10 1093@Zoolinnean@Zlx040

Zoological Journal of the Linnean Society, 2017, XX, 128. With 12 figures.
Hiding in broad daylight: molecular and

morphological data reveal a new ocean sunfish species
(Tetraodontiformes: Molidae) that has eluded recognition
MARIANNE NYEGAARD1*, ETSURO SAWAI2, NEIL GEMMELL3, JOANNE GILLUM3,
NEIL R. LONERAGAN1,4, YUSUKE YAMANOUE5 and ANDREW L. STEWART6
1
School of Veterinary and Life Sciences, Murdoch University, South Street, Murdoch, Western Australia
6150, Australia
2
Graduate School of Biosphere Science, Hiroshima University, 1-4-4 Kagamiyama, Higashi-Hiroshima,
Hiroshima 739-8528, Japan
3
Department of Anatomy, University of Otago, Dunedin 9054, New Zealand
4
Asia Research Centre, Murdoch University, South Street, Murdoch, Western Australia 6150, Australia
5
The University Museum, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
6
Museum of New Zealand, Te Papa Tongarewa, PO Box 467, Wellington 6011, New Zealand
Received 25 November 2016; revised 30 May 2017; accepted for publication 19 June 2017
The taxonomy of the ocean sunfishes (Molidae) has a complicated history. Currently, three genera and four species
are recognized, including two in the genus Mola (M.mola and M.ramsayi). In 2009, a genetic study revealed a poten-
tial third species, Mola speciesC, in Southeast Australian waters. Concentrating on this region, we obtained samples
and morphological data from 27 Mola sp.C specimens, genetically confirmed the existence of this species (mtDNA
D-loop and cytochrome c oxidase 1), and established its morphology across a size spectrum of 50242cm total length.
Mola sp.C is diagnosed by clavus meristics [1517 fin rays (1315 principal, 2 minor), 57 ossicles, paraxial ossicles
separate], clavus morphology (prominent smooth band back-fold, rounded clavus edge with an indent), and body scale
morphology (raised conical midpoints, non-branching). This species does not develop a protruding snout, or swol-
len dorso- or ventrolateral ridges. Body proportions remain similar with growth. Areview of the historic literature
revealed that Mola sp.C is a new, hitherto undescribed species, M.tecta, which we describe and diagnose, and that it
is the first proposed addition to the genus Mola in 125years. Its core distribution is likely in the temperate waters
of the Southern Hemisphere.
ADDITIONAL KEYWORDS: Australia CO1 mtDNA D-loop morphology morphometrics New Zealand
phylogenetics southern hemisphere taxonomy.
INTRODUCTION ranging in life history stages from pre-juveniles (e.g.

Koelreuter, 1766; Pallas, 1770; Mitchill, 1828) to large
The taxonomy of the ocean sunfishes (Molidae
adults (e.g. Ranzani, 1839; Giglioli, 1883; Philippi,
Bonaparte, 1832) has a long and complicated history
1892). The taxonomy quickly became complex due to
dating back more than 250 years, when Mola mola
limited clarity on intraspecific morphological variation,
(Linnaeus, 1758) (originally Tetraodon mola) was first
ontogenetic changes and the potential for sexual dimor-
placed within the Linnaean classification system. Over
phism. This confusion was further compounded by the
the following century, numerous putative new species
difficulty in preserving and storing sub-adult and adult
were described, typically based on single specimens,
specimens. Many specimens were reduced to mounted
*Corresponding author. E-mail: m.nyegaard@murdoch.edu.au skins with varying degrees of accuracy in the reproduc-
[Version of Record, published on 19 July 2017; http:// tion of morphology (e.g. Philippi, 1892; Harting, 1865).
zoobank.org/urn:lsid:zoobank.org:pub:4B1F46B7-CC61- Furthermore, opinions differed as to the correct system-
41FB-BB18-3D1108B9ECF6] atic position of these unusual looking fishes, so species
2017 The Linnean Society of London, Zoological Journal of the Linnean Society, 2017, XX, 128 1
2 M. NYEGAARD ET AL.
were re-grouped and moved between existing and new but recognizing further research is needed to resolve
taxonomic groups, generating a large number of syno- the nomenclature of what appears to be an Atlantic
nyms. Together with spelling deviations, the number and Pacific clade (Ahuir-Baraja, Yamanoue & Kubicek,
of names in the literature for the Molidae soared (e.g. 2017; Sawai etal., 2017).
Fraser-Brunner, 1951) and by the turn of the 19th cen- In addition to the Mola sp. Aand Mola sp. B clades,
tury, the cumulative number of nominal genera and Yoshita etal. (2009) also found evidence of a third Mola
species had surpassed 19 and 50, respectively (Parenti, species (Mola sp. C), based on the D-loop sequence of
2003; Eschmeyer, Fricke & van der Laan, 2017). one specimen from New South Wales in Australia, and
Comparative reviews of the taxonomy were (and still the partial D-loop sequences of two specimens from
are) difficult due to descriptions and drawings with South Africa. The latter two specimens were originally
limited detail and lack of accuracy (e.g. Linnaeus, 1758; from a genetic study by Bass etal. (2005), who inter-
Guilding in Swainson, 1839; Ayres, 1859), the logis- preted them as a M.ramsayi Atlantic group (see Sawai
tics associated with handling, preserving and storing etal., 2017). As no morphological information was col-
the very large adults (e.g. Steenstrup & Ltken, 1898; lected in either study, the taxonomic status of Mola sp.
Matsuura, 2015), and the lack of retention or later loss C remained unresolved.
of holotypes (Parenti, 2003; Eschmeyer et al., 2017). Given the need to formally link genetic clades with
During the 20th century, a general consensus nomenclature, BLAST searches of ocean sunfish DNA
emerged that only a small number of species of sunfish sequences through the Barcode of Life Data Systems
existed. A landmark review of the Molidae by Fraser- (BOLD; www.boldsystems.org) present some pitfalls.
Brunner (1951) synonymized the family to just five spe- Over 100 Molidae mtDNA cytochrome c oxidase 1
cies in three genera: Ranzania Nardo, 1840, Masturus (CO1) sequences are, as of March 2017, lodged with
Gill, 1884 and Mola Koelreuter, 1766. He proposed two BOLD, forming four BINs: Ranzania laevis (Pennant,
species in the genus Mola: M. mola and M. ramsayi 1776) (comprising the majority of the Molidae
(Giglioli, 1883), still recognized by major taxonomic sequences), Masturus lanceolatus (Linard, 1840) and
authorities today (e.g. Bailly, 2015; Eschmeyer et al., two separate clusters of M.mola. Currently, no CO1
2017). Fraser-Brunner (1951) suggested that M. mola sequences labelled M.ramsayi (or Mola spp. A, B or
had a wide distribution in the worlds oceans, but was C) are available; however, blasting a Mola sp. ACO1
replaced by M. ramsayi in the Southern Pacific, with sequence returns a small number of very high similar-
sympatric distributions around Australia. ity scores, pointing to potential mislabelled sequences.
In 2009, Yoshita etal., using phylogenetic analysis This is not surprising, given Mola sp. Ahas been mis-
based on D-loop sequences and accompanying mor- taken for Mola sp. B in various parts of the world
phological data, identified two clades of Mola occur- (Sagara etal., 2005; Yoshita etal., 2009; Thys etal.,
ring sympatrically off Japan. They termed them Mola 2013; Sawai etal., 2017). In light of recent genetic and
group Aand B (hereafter termed Mola sp. Aand Mola taxonomic findings, a review of available Molidae CO1
sp. B), and tentatively linked them to M.ramsayi and sequences would therefore be useful, including to ver-
M.mola, respectively, based on matching morphology ify the existence of Mola sp. C as a separate species.
with Fraser-Brunner (1951). However, one morpholog- During a recent large-scale Molidae biopsy survey
ical discrepancy between Mola sp. Aand M.ramsayi around Australia and New Zealand, we found a subset
could not be resolved: a smooth clavus band (termed of individuals among our material, nesting genetically
band of reduced denticles by Fraser-Brunner, 1951) within the Mola sp. C D-loop clade of Yoshita etal. (2009).
was present on Mola sp. A(Yoshita etal., 2009), whereas The aim of this study is to formally describe Mola sp. C,
M.ramsayi was described by Fraser-Brunner (1951), and this was achieved by first verifying the existence of
and several subsequent authors, as lacking a smooth the clade through analysis of the CO1 locus, comparing
band (e.g. Heemstra, 1986; Glover, 1994; Hutchins, our specimens to Molidae sequences available online.
2001; Bray, 2008; Thys etal., 2013). While further After establishing the common morphological charac-
research is needed to formally resolve this discrepancy, teristics of genetically verified Mola sp. C specimens, we
here we follow Yoshita etal. (2009), and Matsuura reviewed museum material across Australia and New
(2015) in tentatively linking Mola sp. Ato M.ramsayi, Zealand to locate any Mola sp. C specimens held in col-
on the grounds that the smooth band was probably lections. Based on both fresh and preserved specimens,
overlooked on the M.ramsayi holotype, an oversight we describe the morphology of Mola sp. C across a size
subsequently repeated in the literature. Previous and spectrum of 50242cm total length (TL). We compare
recent findings show that Mola specimens from New the morphology to the original descriptions of the syno-
South Wales in Australia, the type locality of M.ram- nyms of M.mola and M.ramsayi to establish whether
sayi, indeed cluster with Mola sp. A(Yoshita etal., Mola sp. C has been described previously. On the basis of
2009; M. Nyegaard, unpublished data). We also follow both genetics and morphology, we conclude that Mola sp.
Yoshita etal. (2009) in linking Mola sp. B to M. mola, C is a new Mola species, which we describe herein.
2017 The Linnean Society of London, Zoological Journal of the Linnean Society, 2017, XX, 128
A NEW OCEAN SUNFISH SPECIES 3
MATERIAL AND METHODS followed by 25 cycles of 94C for 30s, 58C for 40s and
72C for 60s. The PCR products were purified using
Specimen collection
the PALL AcroPrep96 filter plate 30kDa kit prior
A large-scale, non-lethal Molidae biopsy-sampling pro- to Sanger sequencing in forward and reverse using
gram was undertaken in 20132015 in the Australian an ABI Big Dye Terminator v.3.1 cycle sequencing
and New Zealand tuna and billfish long-line fish- kit. Sequencing products were run on an ABI 3730xl
eries, where sunfish are occasionally caught as by- DNA Analyzer through the Genetic Analysis Service
catch. From 2013 onwards, museums across Australia at Otago University, New Zealand. All sequences were
and New Zealand kept us informed of local sunfish uploaded to NCBI with accession numbers provided in
strandings; we obtained photographs, measurements Supporting Information, Table S1.
and genetic samples opportunistically, with the help
of local residents, fisheries department and museum
staff, as well as attending strandings ourselves where Phylogenetic analysis
possible. To acquire specimens for closer examination, D-loop sequences were assembled in McVector
we euthanized three small Mola sp. C from the New v.14.5.3, and pooled with Mola spp. D-loop sequences
Zealand long-line fishery and prepared one of these from Yoshita et al. (2009) (North and Southwest
as holotype at the Museum of New Zealand Te Papa Pacific), Ahuir-Baraja et al. (2017) (Mediterranean)
Tongarewa, Wellington, New Zealand (NMNZ). and Sawai et al. (2017) (Indian Ocean), available from
the National Center for Biotechnology Information
(NCBI). CO1 sequences were pooled with publicly
Review of museum specimens available Molidae CO1 sequences from NCBI and
The Molidae collections at the natural history muse- BOLD (Supporting Information, Table S2). D-loop
ums in Australia and New Zealand were reviewed to and CO1 sequences from one Mola sp. B (261 cm TL)
identify specimens with morphological features akin and four Mola sp. A (24, 58, 120 and 120 cm TL, all
to Mola sp. C.The Melbourne (NMV), Perth (WAM), with smooth clavus bands) from New Zealand were
Auckland (AIM), Otago (OMNZ) and Wellington included as controls, and to confirm the Mola sp. A
(NMNZ) museum collections were inspected in person, clade in the CO1 analysis. To avoid a large number
while material from the Adelaide (SAMA), Darwin of indels in the alignment, which can be problematic
(NTM), Hobart (TMAG), Sydney (AMS), Queensland in phylogenetic inference (Warnow, 2012), we used
(QM), Christchurch (CMC) and Whanganui (WRM) Ma. lanceolatus as the outgroup in the D-loop analy-
museums were inspected via photographs and descrip- sis, rather than the more divergent R. laevis, which
tions provided by museum staff [institutional codes was the outgroup for the CO1 analyses. For both loci,
follow Sabaj (2016), except OMNZ and WRM]. From the pooled sequences were aligned with ClustalW
accessible material, we identified a small number of implemented in MacVector v.14.5.3, using the default
specimens and casts with similar characteristics to settings. The alignments, with a small number of
our genetically confirmed Mola sp. C specimens, and scattered indels for the D-loop, were checked and
morphometric and morphological data were obtained trimmed.
from these. Tissue samples appropriate for genetic For both loci, phylogenetic relationships were
analysis were generally not available from museum inferred using maximum likelihood (ML), Bayesian
specimens (but see Table1). and maximum parsimony (MP) methods, as described
below. The most appropriate model of sequence evolu-
tion was selected based on the corrected Akaike and
Molecular sequencing Bayesian Information Criteria (AICc and BIC), calcu-
Tissue samples were fixed in RNALater or 95% etha- lated in jModelTest v.2.1.10 (Guindon & Gascuel, 2003;
nol. We used a lithium chloride and chloroform salt- Darriba etal., 2012), using the likelihood scores of 88
ing out method (Gemmell & Akiyama, 1996) for candidate models with 11 substitution schemes. This
DNA extraction. The mitochondrial D-loop and CO1 included models with equal/unequal base frequencies
loci were amplified using MolaA and MolaB primers (+F), with/without a proportion of invariable sites (+I)
(Yoshita etal., 2009), and universal fish primers F1 and with/without rate variation among sites (+G) (four
(Ward etal., 2005), respectively. We used Bioline Biotaq rate categories). We used an ML-optimized base tree,
polymerase enzyme for the polymerase chain reaction and a Subtree-Pruning-Regrafting (SPR) tree topol-
(PCR) with each reaction consisting of 0.2L Taq poly- ogy search. The AICc and BIC gave similar results for
merase, 2L 10 buffer, 0.8L MgCl2, 0.4L dNTP the D-loop, with HKY85 (Hasegawa, Kishino & Yano,
(2.5mM each), 0.8L primers, 13.8L water and 1L 1985) +I+G the best fitting model by BIC, and second
template (10ng/L). The 20L reaction mixtures were best fit by AICc. For CO1, HKY+G was the best fitting
subjected to an initial denaturation of 94C for 60s, model based on both BIC andAICc.
Table1. Mola tecta sp. nov. holotype (in bold), paratypes and other material examined as part of this study
Specimen TL (cm) Data Front Clavus Smooth Pectoral Dorsal/ Principal Minor Clavus Paraxial Scale
source profile shape band fin rays anal fin clavus fin clavus fin ossiclesa ossicles shape
back-fold rays raysa rays
NMNZ 49.9 M:m,x Tapered Rounded, Present 12 17/17 15 (7/8) 2 0 Separate Rounded
P.005890 indent (faint)
AIM MA29864 51.1 M:m,x Tapered Rounded, Present 13 19/18 13 (6/7) 2 3 (1/2) Separate Rounded
faint (faint)
indent
NMNZ 57.2 M:m,x Tapered Rounded, Present 12 18/17 14 (6/8) 2 2 (1/1) Separate Rounded
P.033995 indent
NMNZ 60.2 M:m,x Tapered Rounded, Not visible 12 18/17 13 (5/8) 2 2 (0/2) Separate Rounded
P.001418 faint
indent
G06 62 F:g,p Rounded, Present
indent
NMNZ 64.6 M:m Tapered Rounded, Present 11 18/17 3 (1/2) Separate Rounded
P.002980 indent (faint)
NZ18 65 F:g,m,d Tapered Rounded, Present 14 (7/7) 2 6 (3/3) Separate Rounded
indent
(irregu-
lar)
P.006126 indent
NZ19 69 F:g,m,d Tapered Rounded, Presentb 19 (9/10)b 34b 8 (3/5)b Separate Rounded
indent
P.034217 indent
OMNZ VT3249 78 C:m Tapered Rounded, Present Rounded
indent
NMV 78.5 F:p; M:m Tapered Rounded, Present 15 (7/8) 6 (3/3) Separate Rounded
A26565-001 indent
P.034187 indent
NZ17 81 F:g,m Tapered Rounded, Present 14 (6/8) 2 5 (2/3) Separate Rounded
indent
SAMA F7542 Ca. 90 M:g,p Tapered Rounded, Present Rounded
indent
NMV A18725 90.5 M:m Tapered Rounded, Present 6 (2/4) Separate Rounded
indent
NZ07 100 F:g,p Tapered Rounded, Present
indent
Table1. Continued
Specimen TL (cm) Data Front Clavus Smooth Pectoral Dorsal/ Principal Minor Clavus Paraxial Scale
source profile shape band fin rays anal fin clavus fin clavus fin ossiclesa ossicles shape
back-fold rays raysa rays
NMNZ 101.1 F:g,m,x; Tapered Rounded, Present 11 18/17 15 (7/8) 2 7 (3/4) Separate Rounded
P.057679 M:m indent
AIM MA30933 102 C:m,p Tapered Rounded, Present
no
indentc
TMAG D3912 150 F:g,p Rounded, Present
indent
indent
NZ12 155 F:g,m Tapered Rounded, Present 12 13 (6/7) 2 6 (2/4) Separate Rounded
indent
OMNZ 169 F:g,m,p,d Tapered Rounded, Present Rounded
X2017.19 indent
NZ14 170 F:g,p Rounded, Present
indent
indent
NZ01 212 F:g,m,p Tapered Rounded, Present 11
indent
OMNZ VT3248 242 C:m Tapered Rounded, Present 12 Rounded
indent
All specimens: head bump absent; chin bump absent; snout absent; short dorso-lateral ridge present.
TL, total length; C, cast from fresh specimen; F, fresh specimen; M, museum specimen; d, digestive tract content; g, genetic sample; m, morphometric measurements; p, specimen examined from
photographs; x, x-ray.
a
Dorsally/ventrally of the smooth band back-fold.
b
Abnormal clavus, see text for details.
c
Indent likely lost during cast preparation.
A NEW OCEAN SUNFISH SPECIES
5
Maximum likelihood [termed scales herein] between dorsal and anal fins
ML analysis was done in phyML (Guindon & Gascuel, (Fraser-Brunner, 1951).
2003), implemented through the online T-REX server
(Boc, Diallo & Makarenkov, 2012) under the HKY
Clavus and fin meristics
model for both loci. The base frequencies were opti-
mized, and the transition/transversion ratio, gamma We examined the clavus on a subset of fresh and pre-
shape (in four substitution rate categories) and the served specimens (Table1). The number of clavus ossi-
proportion of invariable sites (in the D-loop analysis) cles was established by touch on fresh specimens, and
were estimated by the software. The tree topology by touch and from X-rays, when available, on museum
was optimized with the Best setting [best of SPR and specimens. The number of clavus fin rays, and state
nearest-neighbour interchange (NNI)] from five ran- of paraxial ossicles (Fraser-Brunner, 1951), was estab-
dom BioNJ start trees, and nodal support assessed by lished in fresh specimens after the clavus was cut off
bootstrap from 1000 pseudo-replicates. along the posterior margin of the smooth band, and
dried for several days. The dorsal, anal and clavus fin
ray numbers, and the state of the paraxial ossicles,
Bayesian analysis were determined on museum specimens from X-rays.
Posterior probabilities (PP) for the Bayesian analy-
ses were estimated in MrBayes (Ronquist et al., 2012)
Skin structure
under the HKY model, with rate variation among
sites (in four categories) for both loci, and with a pro- The body scales of the sunfishes are modified into
portion of invariable sites for the D-loop analysis. small spines or denticles (Cleland, 1862; Tyler, 1980;
We used default priors and settings. Two concurrent Gauldie, 1992) and differ between genera in the fam-
analyses were run, each with four Markov chains ily Molidae (Katayama & Matsuura, 2016), as well as
(three heated and one cold) of 10 000 000 (D-loop) between species in the genus Mola (Cadenat, 1959;
and 2 000 000 (CO1) generations, with the chains Gauldie, 1992; Sawai etal., 2015; Sawai, 2016b). We
sampled every 1000 generations. For both loci, con- examined the scale morphology on the body and cla-
vergence between the two concurrent analyses was vus on a subset of fresh, preserved and museum cast
confirmed by checking that the SD of split frequen- specimens (Table1) directly, and through a binocular
cies had decreased below 0.01, and that the effective microscope. Asmall piece of skin was sampled from
sample size was above 200 after a burn-in of 30% two fresh specimens (NZ12 and NZ17) posteriorly of
(Tracer v.1.6; Rambaut et al., 2014). the pectoral fin (Fig.1), dried and compared with ref-
erence skin samples from Mola sp. Aand Mola sp. B
(E.Sawai, unpublisheddata).
MP analysis
For both loci, MP analyses were conducted in Mega v.7
(Kumar, Stecher & Tamura, 2016), including all sites, Morphometric measurements
with an SPR search algorithm (level 1)and five ran- External morphometric measurements follow Yoshita
dom starting trees. Nodal support was estimated using etal. (2009) and a subset of Sawai (2016a) (Fig.1) and
1000 bootstrap pseudo-replicates. were obtained from fresh and preserved specimens, as
well as museum casts (made from fresh specimens),
whenever possible (Table1). The holotype was meas-
External morphology ured both fresh and after the preservation process
The following external characters were assessed in formalin and isopropyl alcohol. We pooled meas-
whenever possible: (1) body proportions: presence or urements from fresh and preserved specimens, as
absence of a head bump (Yoshita etal., 2009; Sawai potential shrinkage from isopropyl alcohol storage is
etal., 2017), and chin bump (Sawai etal., 2017); (2) unlikely to significantly affect the dimensions of pre-
lateral shape of the head profile: presence or absence served specimens (Larochelle etal., 2016). The meas-
of a protruding snout (forward-projecting snout in urements were calculated as a percentage of TL and
Fraser-Brunner, 1951); (3) frontal profile: presence plotted against TL to assess potential change in body
or absence of swollen dorsolateral and ventrolateral shape with size.
ridges on the body (Barnard, 1935); (4) clavus shape:
rounded or lobed/wavy (Fraser-Brunner, 1951; Yoshita
etal., 2009; Sawai etal., 2015), with the clavus being Additional investigations
the movable lobe at the hind margin of the body The digestive tract content of three specimens
(Fraser-Brunner, 1951); and (5) smooth band: present (Table1) was investigated by emptying the content
or absent, defined as the band of reduced denticles into a vessel, separating prey items by hand from the
topographies, with no differences in the relationships

of the major lineages. As expected, they resembled
those of Yoshita etal. (2009), Ahuir-Baraja etal. (2017)
and Sawai etal. (2017) with three major clades: Mola
spp. A, B and C.Our five control sequences from New
Zealand nested within the Mola sp. A(n=4) and Mola
sp. B (n=1) clades, respectively, while the remaining
14 sequences clustered with the Mola sp. C sequence
of Yoshita etal. (2009) (Fig.2). The monophyly of each
Mola clade was well supported with branch supports
of 100% in ML, 1.0 in PP and 100% in MP for each
group. The nodal support for the Mola sp. Aand Mola
sp. C grouping had moderate support (62% in ML and
71% in MP), with the bootstrap results at times plac-
ing Mola sp. Aand Mola sp. B as sister clades.
Details of new CO1 Mola sequences (NCBI acces-
sion numbers MF158115MF158130), and sequences
downloaded from BOLD and NCBI, are provided
in Supporting Information, Tables S1 and S2. The
trimmed alignment of the CO1 sequences comprised
627 characters, with 124 parsimony informative sites.
Figure1. Morphometric measurements following Yoshita Five main clades were produced, each with robust
etal. (2009) and a subset of Sawai (2016a). Body depth
node support of 99100% in ML, 0.951.0 in PP and
(BD), head bump length (HBL), head depth (HD), head
91100% in MP (Fig. 3). Three clades corresponded to
length (HL), pre-anal fin length (PAFL), pre-clavus band
the sequence taxon labels in NCBI/BOLD (i.e. R. lae-
length (PCBL), pre-pectoral fin depth (PPFD), snout length
vis, Ma. lanceolatus and M. mola), while a fourth clade
(SnL), total body depth (TBD), total length (TL). Black box
comprised our new Mola sp. C sequences, ailing with a
indicates region of skin sampling.
BOLD sequence labelled Tetraodontiformes (AMS174-
heavy parasite load and washing prey items in a 600- 08), originating from museum specimen AMS I.41536,
m mesh sieve, as needed, for identification under a registered in OZCAM (2017) as M. mola. However, the
binocular microscope. D-loop from this specimen was included in Yoshita et
al. (2009) as sample NNSW-1, that is Mola sp. C. The
fifth clade comprised our four Mola sp. A sequences
Review of historical records from New Zealand, and one Ma. lanceolatus and two M.
To establish whether Mola sp. C was among previ- mola sequences from NCBI/BOLD. These latter three
ously described Mola species, we reviewed the origi- sequences were probably mislabelled for the following
nal descriptions of M.mola and M.ramsayi, and all reasons: (1) The Ma. lanceolatus sequence KU945271,
nominal Molidae species now considered synonyms of from specimen ASIZP0806237 (Chang et al., 2017), was
M.ramsayi or M.mola, or regarded as nomen nudum originally identified as M. mola in the Fish Database of
by Fraser-Brunner (1951), Parenti (2003) and/or Taiwan (http://fishdb.sinica.edu.tw), indicating poten-
Eschmeyer etal. (2017). In total, 37 species descrip- tial sample or specimen identification uncertainties;
tions were reviewed, along with relevant, associated (2) M. mola sequence AP006238 is the CO1 locus from
literature (Table2). the M. mola mitogenome (Yamanoue et al., 2004), re-
identified as Mola sp. A (sample OI-1) by Yoshita et
al. (2009); and (3) The M. mola sequence HQ167728
originated from a Mediterranean specimen, identified
RESULTS at the time by its 100% similarity with the CO1 locus
of the M. mola mitogenome AP006238, and a lack of
Phylogenetic relationship M. ramsayi records from the area (E. Keskin, Ankara
Details of new Mola D-loop sequences (NCBI acces- University, personal communication, July 2014).
sion numbers MF158131MF158149) are provided in However, mitogenome AP006238 was from Mola sp.
Supporting Information, Table S1. The trimmed align- A (see above), and Mola sp. A has since been recorded
ment of Molidae D-loop sequences (new and from the in the Mediterranean (Ahuir-Baraja et al., 2017). We
literature) comprised 816 characters, with 187 parsi- conclude that the mixed cluster in Fig. 3 is an artefact
mony informative sites. The phylogenetic trees inferred of sequence mislabelling, rather than an indication of
by ML, MP and Bayesian methods displayed similar fundamental problems with the Mola phylogeny. As
Table2. Nominal species listed as Mola sp. or nomen nudum by Fraser-Brunner (1951) (F), Parenti (2003) (P) and
Eschmeyer etal. (2017) (E), with assessment against the Mola tecta sp. nov. morphological characteristics described in
the text
Assessment of nominal species against the Mola tecta morphological characteristics Species identity by taxonomic
authorities
Tetraodon mola Linnaeus, 1758: 334. Type locality: Mediterraneo M.mola (F,P,E)
[Mediterranean Sea]. No types known. Valid species, no citation. Literature source
[Artedi, 1738: 61, 83; Bianchi, 1746: 297; Gronovius, 1754: 55 (No. 125)]. Bianchi (1746)
and Gronovius (1754) indicated Ranzania laevis, but Artedi (1738) indicated Mola sp.
Identification through use of name (see text).
Mola aculeata Koelreuter, 1766: 337, Pl. 8 (figs 2, 3). Type locality: not known. No M.mola (F,P,E)
types known. Original source: two pre-juvenile specimens from the collection at the
Imperial Academy of Sciences in St. Petersburg. Improbable candidate for M.tecta as
specimens were highly unlikely to be from the temperature Southern Hemisphere due
to the limited exploration of these waters at the time of reporting. Attempts to find the
specimens as part of this study were unsuccessful.
Diodon mola Pallas, 1770: 39, Pl. 4 (fig. 7). Type locality: Guinea. No types M.mola (F,P,E)
known. Original source. Unneeded new name (Eschmeyer etal., 2017). Pre-juvenile;
unlikely candidate for M.tecta due to tropical type locality.
Ostracion mola Forsskl, 1775: xviii. Type locality: Malta. Original source. Name nomennudum (P,E)
only, not available (Eschmeyer etal., 2017).
Diodon nummularis Walbaum, 1792: 600. Type locality: not known. No types M.mola (P,E)
known. Literature source [Ruysch, 1710: 26 (No. 63), Pl. 3 (fig.7); Koelreuter, 1766:
337, Pl. 8 (figs 2, 3); Pallas, 1770: 39, Pl. 4 (fig.7); Gmelin, 1788: 1452a (No. 3)]. Ruyschs
(1710) description was superficial, based on a pre-juvenile specimen presumably from
Indian waters (given as the old Dutch Indiaanisch and Latin Indicus). Gmelin (1788)
was based on literature sources, including Pallas [1770: 39, Pl. 4 (fig.7)], and Gronovius
[1754: 55 (No. 125)]. The latter was based on literature sources Artedi (1738: 61)and
Jonstonus [1657: Pl. 9 (No. 1)], both describing adult forms of Mola sp. and R.laevis,
respectively. Diodon nummularis is an unlikely candidate for M.tecta due to the locali-
ties, and confusion of literature sources.
Mola rotunda Cuvier, 1798: 324. Type locality: France. No types known. M.mola (F,P,E)
Presumably literature source (unknown). Unneeded new name for T.mola Linnaeus,
1758 (Eschmeyer etal., 2017).
Orthragoriscus fasciatus Bloch & Schneider, 1801: 511. Type locality: mari sep- M.mola (F,P,E)
tentrionali [northern ocean; presumably the North Sea]. No types known.
Literature source (Duhamel du Monceau, 1777: 306, Pl. 23). Illustration apparently of a
specimen from Angola, Africa, but presumably sufficiently similar to include as illustra-
tion of O.fasciatus. Unlikely candidate for M.tecta due to description and illustration of
protruding snout.
Orthragoriscus hispidus Bloch & Schneider, 1801: 511. Type locality: not known. M.mola (F,P,E)
No types known. Literature source [Pallas, 1770: 39, Pl. 4 (fig.7); Gmelin, 1788:1452].
Pre-juvenile; unlikely candidate (see D.nummularis).
Cephalus brevis Shaw, 1804: 437, Pl. 175 (top). Type locality: European seas. No M.mola (F,P,E)
types known. Literature source (Willughby, 1686: 151, Pl. I.26; Linnaeus, 1758: 334;
Pennant, 1776: 131; Bloch, 1785: 75, Pl. 125). Unneeded new name based on previously
described species and literature sources (Eschmeyer etal., 2017). Unlikely candidate
for M.tecta due to mention of commonness in the Mediterranean (Bloch, 1785; Pennant,
1776).
Cephalus pallasianus Shaw, 1804: 440. Type locality: tropical seas. No types M.mola (F,P,E)
known. Literature source [Pallas, 1770: 39, Pl. 4 (fig.7); Gmelin, 1788:1452]. Pre-
juvenile, unlikely candidate for M.tecta (see D.nummularis).
Orthragus luna Rafinesque, 1810a: 17. Type locality: Sicilia [Sicily, i.e. M.mola (F,P,E)
Mediterranean Sea]. No types known. Literature source (Linnaeus, 1758: 334). We
consider this an unneeded new name for T.mola Linnaeus, 1758.
Table2. Continued
authorities
Orthragus commersoni Rafinesque, 1810a: 17. Type locality: not known. Unknown R.laevis (F), nomennudum
source. Name only, unavailable (Eschmeyer etal., 2017). Unlikely candidate for M.tecta (P,E)
due to context of Rafinesque (1810a) indicating R.laevis.
Diplanchias nasus Rafinesque, 1810a: 17. Type locality: Sicilia [Sicily, i.e. M.mola (F,P,E)
Mediterranean Sea]. No types known. Original source. Also in Rafinesque (1810b:
40)as Diplanchias mola (Eschmeyer etal., 2017). Unlikely candidate for M.tecta due to
locality and description of protruding snout.
Mola aspera Nardo, 1827a: 26, 35 (No.26). Type locality: Adriatica [Adriatic Sea]. ?M.mola (F), M.mola (P,E)
No types known. Literature source [Gronovius, 1763: 50 (No. 186); Bloch, 1786: 83, Pl.
128; Daubenton, 1787: 240 (LUNE)]. Fraser-Brunner (1951: 115)and Eschmeyer etal.
(2017) treated this as a questionably synonym. Insufficient taxonomic characters for
assessment, however unlikely candidate for M.tecta due to locality.
Mola hispida Nardo, 1827b: 104. Type locality: Adriatico mari [Adriatic Sea]. No M.mola (F,P,E)
types known. Literature source [Koelreuter, 1766: 337b, Pl. 8c (figs 2, 3); Pallas, 1770:
39, Pl. 4 (fig.7d); Gmelin, 1788:1452]. Subjectively invalid, secondarily preoccupied in
Mola by O.hispidus Bloch & Schneider, 1801 (Eschmeyer etal., 2017). Pre-juvenile,
unlikely candidate for M.tecta (see D.nummularis).
Cephalus ortagoriscus Risso, 1826: 173. Type locality: Nice, France, i.e. M.mola (F,P,E)
Mediterranean Sea. No types known. Original source (Risso, 1810: 60)and litera-
ture sourcee (Rondelet, 1558: 326; Bloch, 1786: 83, Pl. 128, Cuvier, 1817: 148). Unlikely
candidate for M.tecta due to the number of clavus fin rays (18), its commonness in the
Mediterranean, and the mention of sexual dimorphism.
Diodon carinatus Mitchill, 1828: 264, Pl. 5 (fig. 1). Type locality: Massachusetts, M.mola (F,P,E)
USA. No types known. Original source. Pre-juvenile, unlikely candidate for M.tecta
due to type locality.
Pedalion gigas Guilding in Swainson, 1838: 199, fig.33. Type locality: West M.mola (F,P,E)
Indies. No types known. Original source. Unlikely candidate for M.tecta due to wavy
clavus on illustration.
Molacanthus pallasii Swainson, 1839: 329. Type locality: not known [given as M.mola (F,P,E)
Guinea in Pallas, 1770]. No types known. Literature source [Pallas, 1770: 39, Pl. 4
(fig.7)]. We consider this an unneeded new name based on previously described species.
Pre-juvenile, unlikely candidate for M.tecta (see D.mola).
Tympanomium planci Ranzani, 1839: 74, 76, table. Type locality: mari adriatico M.mola (F,P,E)
[Adriatic Sea]. Holotype (unique): whereabouts unknown. Literature source
(Bianchi, 1755: 331). Subjectively invalid, preoccupied in Mola by M.planci Nardo
(1827a) (Eschmeyer etal., 2017). Unlikely candidate for M.tecta due to protruding
snout and scalloped clavus.
Trematopsis willugbei Ranzani, 1839: 74, table. Type locality: oceano [ocean] M.mola (F,P,E)
[given as Mediterraneo, atque etiam in Oceano (Mediterranean and also
in the Ocean) in Willughby, 1686]. Holotype (unique): not found at MZUB.
Literature source (Willughby, 1686: 151, Pl. I.26), based on original source. We consider
the validity of the species doubtful, as it was based on four holes in a rectangle on top
of the head (Willughby, 1686), likely from external damage. Unlikely candidate for
M.tecta due to locality, description of a rounded clavus without mention of an indenta-
tion, and a protruding snout on the included illustration by Salviani (1554: 154).
Ozodura orsini Ranzani, 1839: 75, 80, table, Pl. 6 (right). Type locality: mari M.mola (F,P,E)
adriatico [Adriatic Sea]. Holotype (unique): mount on wall at MZUB. Extant
at MZUB, no registration number available. Original source. Unlikely candidate for
M.tecta due to locality, description and illustration, in particular a rounded clavus
without an indentation, and merged paraxial ossicles.
Table2. Continued
authorities
Orthragoriscus retzii Ranzani, 1839: 75, table. Type locality: not known [given M.mola (F,P,E)
as Landscrona (i.e resund between Denmark Sweden) in Retzius, 1785]. No
types known. Literature source (Retzius, 1785: 115, Pl. 4; Retzius, 1800: 310f). Based
on Retzius (1785) (Eschmeyer etal., 2017). Unlikely candidate for M.tecta due to loca-
tion, rough skin and detailed description and illustration of a prominent smooth band
without mention of a back-fold.
Orthragoriscus ghini Ranzani, 1839: 75, table. Type locality: mari mediterraneo M.mola (F,P,E)
[Mediterranean Sea]. No types known. Literature source (Salviani, 1554: 154).
Unlikely candidate for M.tecta due to protruding snout on illustration (Salviani, 1554:
154).
Orthragoriscus rondeletii Ranzani, 1839: 75, table. Type locality: mari mediter- M.mola (F,P,E)
raneo [Mediterranean Sea]. No types known. Literature source (Rondelet, 1554:
424). Lack of taxonomic details precludes assessment, but unlikely candidate for
M.tecta due to commonness at the locality.
Orthragoriscus redi Ranzani, 1839: 75, table. Type locality: mari mediterraneo M.mola (F,P,E)
[Mediterranean Sea]. No types known. Literature source (Redi, 1684: 166). Vague
description; lack of taxonomic details precludes assessment. Unlikely candidate for
M.tecta due to commonness at the locality.
Orthragoriscus alexandrini Ranzani, 1839: 75, 78, table, Pl. 6 (left). Type locality: M.mola (F,P,E)
mari adriatico [Adriatic Sea]. No known types. Original source. Unlikely candi-
date for M.tecta due to description and illustration of large head bump and rounded
clavus without indentation.
Orthragoriscus blochii Ranzani, 1839: 76, table. Type locality: mari oceano M.mola (F,P,E)
[ocean sea, possibly Atlantic]. No types known. Literature source [Bloch,
1786: 83, Pl. 128; Bonnaterre, 1788: 54, Pl. 17 (fig.54); Shaw, 1804: 437, Pl. 175 (top)].
Unlikely candidate for M.tecta due to description of rough skin, and mention of com-
monness at the locality.
Orthragoriscus elegans Ranzani, 1839: 76, table. Type locality: mari atlantico R.laevis (F), M.mola (P,E)
[Atlantic Sea]. No types known. Literature source (Bloch & Schneider, 1801: 511,
Pl. 97). Parenti (2003: 3)and Eschmeyer etal. (2017) synonymized O.elegans with
M.mola. However, O.elegans was based on O.oblongus in Bloch & Schneider (1801:
511, Pl. 97), which is considered a synonym of R.laevis by Parenti (2003: 4)and
Eschmeyer etal. (2017). Unlikely candidate for M.tecta as the illustration in Bloch &
Schneider (1801: Pl. 97)is clearly R.laevis.
Pallasia pallasi Nardo, 1840: 112. Type locality: not known [given as Guinea in M.mola (F), nomennudum
Pallas, 1770]. Literature source [Pallas, 1770: 39, Pl. 4 (fig.7)]. Pallasina pallasi [sic] (P,E)
in Fraser-Brunner (1951: 110, 115), name deviation discussed in Parenti (2003: 6)and
Eschmeyer etal. (2017). Unneeded new name based on previously described species.
Pre-juvenile; unlikely candidate for M.tecta (see D.mola).
Orthragoriscus lunaris Gronow in Gray, 1854: 165. Type locality: Mediterranean R.laevis (F), M.mola (P,E)
Sea. No types known. Literature sourceg [Boussuet, 1558: 179; Gronovius, 1763: 50
(No. 185); Linnaeus, 1766: 412 (gen. 137, No 7); Bianchi, 1755: 331; Borlase, 1758: 268,
Pl. 26 (fig.7); Klein, 1742: 23 (No. 31)]. This author probably meant to describe a Mola
species but confused the characteristics and references of R.laevis and Mola sp., pre-
cluding assessment.
Orthagoriscus analis Ayres, 1859: 31, fig. 5 (p.54). Type locality: Santa Barbara M.mola (F,P,E)
Channel, California, USA. Holotype (unique): CAS (lost in 1906). The holotype
was unregistered, and lost along with all records in an earth quake and subsequent
fire (D. Catalina, California Academy of Sciences, personal communication, April 2017).
Original source. Specimen with damaged dorsal and anal fins (Gnther, 1870: 319).
Description rudimentary, illustration extremely crude, precludes identification below
genus level. Unlikely candidate for M.tecta based on locality.
Table2. Continued
authorities
Aledon storeri Castelnau, 1861: 76. Type locality: not known [given as Boston M.mola (F,P,E)
Bay, Massachusetts, USA.in Storer, 1839]. No types known. Literature source
[Storer, 1839: 515, Pl. 3 (fig.1)]. Storer (1839) based on original source. Unlikely candi-
date for M.tecta based on scalloped clavus and a protruding snout.
Pedalion capensis Castelnau, 1861: 75. Type locality: la baie dela Table [Table M.mola (P,E)
Bay, South Africa]. No types known. Original source. Unlikely candidate for
M.tecta due to description of rough skin, protruding snout and scalloped clavus.
Aledon capensis Castelnau, 1861: 75. Type locality: la baie dela Table [Table Bay, M.mola (F,P)
South Africa]. New combination for P.capensis described by Castelnau in the same
paper, placed into new genus Aledon at the end of the article (Eschmeyer etal., 2017);
an unneeded new name.
Orthragoriscus ramsayi Giglioli, 1883: 315. Type locality: New South Wales, M.ramsayi (F,P,E)
Australia. Type BMNH 1883.11.29.22; stuffed skin, poor condition; restored
in 2017 (see text). Holotype listed as BMNH 1888.11.29.22 in Eschemeyer (2017).
Original source. Unlikely candidate for M.tecta (see text).
Orthagoriscus eurypterus Philippi, 1892: 14, Pl. 6 (fig. 1). Type locality: Chaaral, ?M. ramsayi (F), M.ramsayi
Chile. Mounted skin at MNHNCH, unregistered, not designation as holotype (P,E)
(see text). Original source. Unlikely candidate for M.tecta (see text).
Original source means the species description was based on a specimen; literature source means the species description was based on existing
literature. Type status from Eschmeyer etal. (2017) unless otherwise stated. Museum codes according to Sabaj (2016).
a
Page1450 in Walbaum (1792), presumably a typographical error.
b
Page 440 in Nardo (1827b); cPl. 6 in Nardo (1827b); dfig. 9 in Nardo (1827b) presumably typographical errors.
e
We could not resolve referenced work: Salv., 55 (possibly Salviani, 1554: 154); Will., 16. (possibly Willughby, 1686: 151. Pl. I.26); Lac., 2,364 (possibly
Lacpde, 1798: 633).
f
Page 320 in Ranzani (1839): table.
g
We could not resolve referenced work: Charlet, Onom. Zoic. Pisc. P.9$4 [possibly Charleton, 1668: 129 (No. 3)].
such, the CO1 analysis corroborates the D-loop results, openings small, oval, in front of pectorals. Lateral lines
with three clear Mola species clades, Mola sp. B (M. around eyes (nearly imperceptible). No swim bladder.
mola), Mola sp. A (presumably M. ramsay) and Mola Darker above, paler or dusky below, often spots and
sp. C, showing the latter two as sister clades. patterns on the sides.
Mola mola (Linnaeus,1758)

DESCRIPTION
Mola mola (Linnaeus, 1758). No known holotype
Family Molidae Bonaparte, 1832 (Parenti, 2003; Eschmeyer et al., 2017)
Genus Mola Koelreuter,1766 Distribution: Currently thought to be widely distrib-
Mola Koelreuter, 1766. Type specimen Mola aculeata, uted in all the worlds oceans except for the polar seas,
i.e. pre-juvenileform. however may be relatively uncommon in some areas
Mola Cuvier, 1798. Type specimen Tetraodon mola (e.g. around Australia and New Zealand).
Linnaeus, 1758, i.e. adultform.
Diagnosis: Genus diagnosis, modified from Parenti Mola ramsayi (Giglioli,1883)

(2003): Body ovoid, highly compressed laterally; lacking
Mola ramsayi (Giglioli, 1883). Holotype: BMNH
a true caudal fin; caudal region with a pseudocaudal
1883.11.29.22
(clavus), rounded, sometimes slightly crenulated, but
without a mid-flap extension. Mouth small; teeth Distribution: Previously thought to be restricted to the
united, fused and beaklike. Fins without spines; dorsal South Pacific (Fraser-Brunner, 1951), however if Mola
fin posterior, erect, high, short-based, anal fin opposite sp. A is shown to belong to M. ramsayi, the distribution
the same shape. Pelvic fins absent. Skin leathery and is much wider than previously thought, including the
thick; scales reduced, rounded base with a single, Pacific and Indian oceans in both hemispheres, as well as
erect point or rectangular shape; non-imbricated. Gill the Mediterranean and probably also the European seas.
Figure 2. Phylogenetic relationships inferred by maximum likelihood (ML), based on Mola mitochondrial D-loop sequences
from Yoshita et al. (2009), Ahuir-Baraja et al. (2017), Sawai et al. (2017) and this study (in bold). Numbers above or below
internal branches of the major clades of Mola sp. A, Mola sp. B, and M. tecta (Mola sp. C) refer to ML bootstrap values (left),
Bayesian posterior probabilities (middle) and maximum parsimony bootstrap values (right). Scale bar represents nucleotide
substitution.
Figure3. Phylogenetic relationships inferred by maximum likelihood (ML), based on Molidae mitochondrial cytochrome
c oxidase 1 sequences from the Barcode of Life Data Systems, the National Center for Biotechnology Information and this
study (in bold). Numbers above or below internal branches of the major species clades refer to ML bootstrap values (left),
Bayesian posterior probabilities (middle) and maximum parsimony bootstrap values (right). Scale bar represents nucleotide
substitution. *Lodged as Tetraodontiformes; from specimen AMS I.41536, also NNSW-1 in Yoshita etal. (2009), i.e. Mola sp.
C.**Probably Mola sp. A(see text). ^Sample OI-1 in Yoshita etal. (2009), i.e. Mola sp. A.^^Probably Mola sp. A(see text);
*^Lodged as Tetraodontiformes, from specimen NMV A25071, Mola sp. in OZCAM (2017).
Mola tecta sp. nov Paratypes: Eleven specimens (49.990.5 TL): AIM
(Table 1; Figs 110,12) MA29864 (51.1 cm TL), female, whole specimen
in isopropyl alcohol, Poor Knights Islands, New
urn:lsid:zoobank.org:act:0AE4167C-A2FE-4E77-
Zealand (3529.3'S 17343.7E), 8 November 1969;
BD45-B0D465410988
NMNZ P.001418 (60.2 cm TL), male, whole specimen
New English name: Hoodwinker ocean sunfish. New in isopropyl alcohol, off Hawkes Bay, New Zealand
Japanese name: Kakure-manbo (3925S 1776E), June 1952; NMNZ P.002980
(64.6 cm TL), male, whole specimen in isopropyl
Orthragoriscus mola (not of Linnaeus) Lidth de
alcohol, North Rona Bay, Eastbourne, Wellington
Jeude, 1890: 189190, Pl. 8.
Harbour, New Zealand (4116S 17455E), drag
Mola ramsayi (not of Giglioli) Gauldie, 1992:
net, coll.: Robert Falla, 30 November 1960; NMNZ
263266 [description of skin structure], figs 46
P.005890 (49.9 cm TL), female, whole specimen
Bass et al., 2005: 405413 [as Atlantic group],
in isopropyl alcohol, off Great Barrier Island, New
figs 2, 3, table 1; accession numbers AY940816
Zealand (365S 17535E), trawl 3637 m, coll.: C.
and AY940826 (partial D-loop); AY940834 and
Gaelic, 25 June 1973; NMNZ P.006126 (64.8 cm
AY940838 (cyt-b).
TL), female, whole specimen in isopropyl alcohol,
Mola group C Yoshita etal., 2009: 237, NNSW-1 in
Oriental Bay, Wellington Harbour, New Zealand
table1 and figs 3, 4; Accession number AB439108
(4117.40S 17447.587E), beach cast, 29 June 1955;
(D-loop).
Mola species C of Yoshita etal., 2009 Ahuir-Baraja
in isopropyl alcohol, Bay of Plenty, Opotiki Beach,
etal., 2017: 1133, table SI, fig.2 Sawai etal., 2017:
New Zealand (3745S 17720E), November 1996;
99102, figs 1, 2, table1.
in isopropyl alcohol, Bay of Plenty, surf at Opotiki,
H o l o t y p e : N M N Z P. 0 5 7 6 7 9 ( F i g .4 ) ( f r e s h New Zealand (3745S 17720E), hand, coll.: Andy
measurements: 101.1cm TL; measurements after Glazier; NMNZ P.034217 (69.8 cm TL), male, whole
fixation: 101.2cm TL), male, North Taranaki Bight, specimen in isopropyl alcohol, Bay of Plenty, Opotiki
west coast North Island, New Zealand (3825.5S Beach, New Zealand (3744S 17720E), coll.: Andy
1749.0E), trawl, 78 m, 25 December 2015; coll.: Glazier, December 1996; NMV A18725 (90.5 cm TL),
observer Scott Yeoman FV Ivan Golubets, OBS whole specimen in isopropyl alcohol, Port Phillip
4552/050. Measurements (fresh and after fixation) Bay, Victoria (3752S 14449E), coll.: F. Hadathy &
provided in Supporting Information, TableS3. S. Praljak, 28 August 1995; NMV A26565-001 (78.5
Figure4. Mola tecta sp. nov. holotype: NMNZ P.057679, 101.1cm total length (fresh measurement), male, North
Taranaki Bight, west coast North Island, New Zealand (3825.5S 1749.0E), trawl, 78 m, 25 December 2015, collected by
Scot Yeoman, Ministry of Primary Industries, New Zealand. (A) Photographed and (B) illustrated by Carl Struthers and
Michelle Freeborn, respectively, Museum of New Zealand Te Papa Tongarewa.
cm TL), whole specimen in isopropyl alcohol, Barwon characters: slender body shape without head bump
Heads, Victoria (3817.4E 14427.36S), coll.: Warren or chin bump; tapered profile when viewed from the
Chapman, 30 April 2009; SAMA F7542 (c. 90 cm front, with the widest point across the eye area; short
TL), whole specimen in ethanol, Spencer gulf, South dorsolateral ridge to posterior end of pectoral fin;
Australia (3447.4S 13828.8E), coll.: J. Verrier, 27 rounded profile without protruding snout. Pronounced
June 1994 [current specimen label indicates 1989, smooth band consisting of creased, elastic skin with
considered an error by collection authority (R. Foster, highly reduced density of body scales; distinctive
SAMA, personal communication, February 2017)]. smooth band back-fold diving the clavus into a smaller
upper and larger lower portion. Rounded clavus,
Other material: Sixteen specimens (62242 cm TL): slightly crenulated, with an indentation associated with
AIM MA30933 (102 cm TL), cast from fresh specimen, the smooth band backfold. Usually 1113 pectoral fin
New Zealand; G06 (62 cm TL), live specimen (not rays, 1719 dorsal fin rays, 1618 anal fin rays, 1517
retained), northeast of North Island, New Zealand, clavus fin rays (1315 principal rays, 2 minor rays
(3630S 17745E), long-line, 8 July 2013; NZ01 (212 along the edges of the smooth band back-fold); usually
cm TL), fresh specimen (not retained), Otago Harbour, 57 separate ossicles on the clavus edge in specimens
New Zealand (4549S 17037E), beach cast, 18 over 65cm TL; no ossicles or ossification associated
February 2015; NZ07 (100 cm TL), fresh specimen (not with the smooth band back-fold, ossicles on paraxial fin
retained), Banks Peninsula, New Zealand (4350S rays separate, of similar size to the neighbouring clavus
17240E), beach cast, 30 April 2014; NZ08 (151 cm ossicles. Body scales with conical, non-branching erect
TL), fresh specimen (not retained), Banks Peninsula, central point of differing sizes.
New Zealand (4350S 17240E), beach cast, 30
April 2014; NZ09 (193 cm TL), fresh specimen (not Description: Description of the holotype (Fig.4)
retained), Banks Peninsula, New Zealand (4350S followed by a range of paratypes and sample
17240E), beach cast, 30 April 2014; NZ12 (155 cm s p e c i m e n s i n b r a c k e t s ( s e e Ta b l e 1 f o r t h e
TL), female, fresh specimen (clavus, skin sample contribution of data from individual paratypes and
retained), Birdlings Flat, Banks Peninsula, New sample specimens).
Zealand (4349S 17242E), beach cast, 10 May 2014; A species of Mola with the typical characteristics
NZ14 (170 cm TL), fresh specimen (not retained), for the genus, most notably a deep body, highly com-
near Birdlings Flat, Banks Peninsula, New Zealand pressed laterally, truncated, lacking a true caudal fin,
(4350S 17240E), beach cast, 14 December 2015; near-symmetrical, spineless dorsal and anal fins. Eyes
NZ17 (81 cm TL), male, fresh specimen (clavus, skin small, round; two small nostrils located in front of eye;
sample retained), west of South Island, New Zealand mouth small, terminal; teeth fused, beaklike; gill open-
(4132S 17056E), purse seine, 12 April 2014; NZ18 ings small, oval, located in front of the pectoral fins,
(65 cm TL), female, fresh specimen (clavus retained), and covered by a soft gill membrane; gill rakers are
east of North Island, New Zealand (3847S 17825E), concealed under the subcutaneous gelatinous layer;
long-line, 17 May 2014; NZ19 (69 cm TL), male, fresh all fins are spineless; caudal fin and pelvic fin absent;
specimen (clavus retained), east of North Island, New pectoral fin small, round, located mid-laterally, fitting
Zealand (359S 1763E), long-line, 10 August 2014; into a shallow groove in the side of the body; caudal
OMNZ VT3249 (78 cm TL), cast from fresh specimen, fin replaced by a clavus, comprised of highly modified
Kaka Point, Clutha District, New Zealand, (4622S elements of the dorsal and anal fins; body covered in
16944E), beach cast, 7 March 1963; OMNZ VT3248 thick subcutaneous gelatinous layer except for pecto-
(242 cm TL), cast from fresh specimen, Otago Harbour, ral, dorsal and anal fins, skin covered in thick mucus,
Dunedin, New Zealand (4549S 17038E), beach cast, anal opening immediately anterior of analfin.
1961; OMNZ X2017.19 (169 cm TL), female, fresh No head bump (0), no chin bump (0), no swollen dor-
specimen (not retained), North Taranaki Bight, New solateral or ventrolateral ridges on body (0); short dor-
Zealand (3826S 1749E), beach cast, 18 January 2017; solateral ridge to posterior end of pectoral fin reflecting
RMNH. PISC.D.2059 (223 cm TL), female, mounted the underlying cranio-skeletal structure, front profile
skin, Ameland, Netherlands, beach cast, 13 December tapered (in all investigated specimens) (Figs 5, 6).
1889; TMAG D3912 (150 cm TL), fresh specimen (not Lateral profile rounded, without protruding snout
retained), Lindisfarne, Tasmania, Australia (4251S (0) or any indications thereof (0) (Fig.5B). Holotype
14720E), beach cast, 12 December 2014. of typical body dimensions (Fig.7); M.tecta maintains
its body proportions with growth, with similar length
Diagnosis: Within the investigated size range (50 and depth measurements as percentages of TL across
242cm TL), M.tecta sp. nov. is diagnosed from others the investigated size spectrum; total body depth (TBD)
in the genus Mola by the following combination of relative to TL appears to decrease with growth, due
Figure5. Mola tecta sp. nov. specimen NZ12, 155cm total length. (A) Tapered profile (hashed lines) with widest point
across the eye area; (B) rounded profile without protruding snout (black arrow); short dorsolateral ridge (white arrows) to
pectoral fin; (C) prominent smooth clavus band with back-fold (black arrows), and associated indent (white triangle) on an
otherwise rounded clavus edge. White stars indicate ossicles.
to shorter dorsal and anal fins relative to TL in larger below the smooth band back-fold. An additional 2 (2)
specimens (Figs 6, 7A). minor fin rays embedded along the edges of the smooth
Eleven (1113) pectoral fin rays, 18 (1718) dorsal band back-fold; minor fin rays and back-fold positioned
and 17 (1618) anal fin rays, 15 (1315) regularly spaced between the two paraxial fin rays (Fig. 8). Seven (07)
principal clavus fin rays, with 7 (57) above, and 8 (78) ossicles: 3 (03) above, and 4 (04) below the smooth
Figure6. Slender body proportions without head bump, chin bump or protruding snout of Mola tecta sp. nov. at differ-
ent total lengths: (A, G) G06: 62cm; (B) NZ18: 65cm; (C) NZ07: 100cm; (D) NZ08: 151cm; (E) NZ01: 212cm; (F, I) OMNZ
VT3248 242cm (cast); (H) NMNZ P.057679 (holotype): 101cm. Photographs by Tom Trnski, Auckland War Memorial
Museum (A), Marianne Nyegaard, Murdoch University (B, F), Ken Logan, local resident of Christchurch, New Zealand
(C, D), Jean McKinnon, University of Otago (E). Illustrations by Michelle Freeborn, Museum of New Zealand Te Papa
Tongarewa (GI).
band back-fold (Figs 4, 5C), positioned on the crest of sizes, on the elastic skin of the smooth band back-fold
slight crenulations on the clavus trailing edge (Fig. (density of scales on back-fold varies between individ-
8B,C), adult ossicle number established around 65 cm. uals). Clavus rounded with an indent at smooth band
Paraxial ossicles separate, situated on either side of the back-fold (Figs 4, 5C) (back-fold and indent may be less
smooth band back-fold (in all investigated specimens) evident in small, preserved specimens due to stiffening
(Fig. 8). of the back-fold during preservation).
Shape of scales is similar across much of the body and Coloration of holotype after defrosting uniform grey,
clavus. For comparison with other Molidae, we refer darker dorsally, lighter ventrally (Fig.4). The colour of
here to the skin on the body posterior of the pectoral live specimens appears grey or dark brown with dusky
fin (Fig.1). Scales with a raised central point, which white, mottled spots and patterns (Fig.6A; Supporting
is smooth and conical without branching tips (Fig.9). Information, Fig. S1). Darker dorsally, lighter ven-
The scales butt against each other; the boundaries trally. It is not known whether live M.tecta can rapidly
become highly visible in the dried state (not shown). change the contrast of its skin patterns, as Mola sp.
Size of the raised, central point on the body scales vary, Aand Mola sp. B can (Nyegaard M and Sawai E, per-
with larger scattered between smaller (Fig.9A, B). sonal observation), but we consider it likely. The colour
Smooth band nearly free of scales (in all investigated of freshly dead specimens is similar to live specimens,
specimens); scales only slightly reduced, and of smaller or uniform grey to grey-blue (Figs 5, 6BE).
the English name, Hoodwinker, pertains to the

figurative meaning trickster, deceiving by disguise,
c. 1600.
Habitat and distribution: Mola tecta has been

confirmed in the southeast of Australia (New South
Wales, South Australia, Tasmania and Victoria)
(Yoshita etal., 2009; this study), around New Zealand
(this study), and off South Africa (Bass etal., 2005)
(Fig.10). Photographs (e.g. Supporting Information,
Fig. S1A) indicates that it also occurs in Chilean
waters. This suggests that M.tecta is distributed
widely in the temperate waters of the Southern
Hemisphere, with occasional, but presumably rare,
occurrences in the Northern Hemisphere (see below
under Broader Molidae literature). Migrations are
unknown.
Biology: It attains at least 242cm TL, probably more.

It likely exhibits a similar biology to the other Mola
species whereby feeding takes place during deep dives
(e.g. Nakamura, Goto & Sato, 2015; Thys etal., 2016).
Digestive tract contents of three specimens (Table1)
consisted of salps (Thetys vagina and Pyrosoma sp.),
and the remains of a nectonic siphonophore, and in
one instance, a 35mm Styrofoam ball. All dissected
specimens were heavily infested with parasites,
particularly in the intestines (cestodes) and in the liver
(likely larval Trypanorhynch cestodes). The shape of the
gonads differs between the sexes; the ovary is singular
and ball-shaped, the testis are paired, elongated and
rod-like and do not appear to frill during maturation.
Spawning grounds, eggs, larvae and pre-juveniles are
not known.
Remarks: The clavus fin ray structure of NZ19

(Table1) appeared abnormal with a secondary,
Figure7. Morphometric measurements of Mola tecta sp. irregular smooth band back-fold present below the
nov. specimens (Table1), as percentage of total length (TL). main back-fold. Also, in addition to 14 regularly
(A) Black diamonds: body depth (BD), black squares: head spaced principal fin rays, we found 5 smaller, less rigid
depth (HD), open triangles: pre-pectoral fin depth (PPFD), and highly irregular fin ray-like structures; 4 of these
open circles: total body depth (TBD); (B) open diamonds: were set close to each other, bending towards the main
head bump length (HBL), open triangles: head length (HL), back-fold, while the fifth was Y-shaped and associated
black diamonds: pre-anal fin length (PAFL), open circles: with the secondary back-fold. Their irregularity and
pre-clavus band length (PCBL), black squares: snout length disordered appearance pointed to malformations,
(SnL). Holotype NMNZ P.057679 morphometrics (fresh) which are not uncommon in the other Mola species
indicated with arrows.
(Sawai etal., 2009). The formation of ossicles (8) also
appeared affected.
Etymology: The species name tecta is derived from
the Latin tectus (disguised, hidden), as this species
evaded discovery for nearly three centuries, despite
the keen interest among early sunfish taxonomists Key to the Mola species
and the continued attention these curious fish receive. Fraser-Brunners (1951) review included a key to the
The Japanese name is derived from the species name genus Mola, whereby M. mola and M. ramsayi were
tecta: hidden (Kakure), sunfish (manbo), while distinguished based on six characters: the number of
Figure 8. Clavus structures of Mola tecta sp. nov. specimen NZ12, 155 cm total length. (A) Dried clavus, left side,
skin partially removed: two minor fin rays (black arrows) embedded along the edge of the smooth band back-fold (white
diamond) between the paraxial fin rays (white arrows), which bear separate ossicles (white triangles). (B) Fresh and (C)
dried state, right side: small ossicles (triangles) on the crests of slight crenulations along the rounded clavus edge; no
ossicles or ossification associated with the clavus band back-fold (white diamond); paraxial fin rays (white arrows) each
bearing a separate ossicle (white triangles), similar to the ossicles (black triangles) on neighbouring principle fin rays
(black arrows).
Figure9. Scales of Mola teca sp. nov. (holotype NMNZ P.057679, fresh state) on the body behind the pectoral fin (see
Fig.1) from (A) a 90degree angle, and (B) an oblique angle (1cm scale bars), with conical, non-branching central points,
larger (black arrows) scattered among smaller (white arrows) scales. (C) Typical shape of Mola tecta sp. nov. body scale:
side view (left), and 90degree angle (right), showing the raised central point on the dermal plate.
clavus fin rays and ossicles, the size of clavus ossicles Mola key. Further research is also needed to determine
against the space between them, the clavus shape (lobed at what size specific characters are established, for
or rounded), state of the paraxial ossicles (separate or example the number of ossicles, wavy clavus, etc. Thus,
united) and presence/absence of a smooth band. While the interim key below should be applied with caution,
research in recent years have added important mor- in particular to smaller individuals. Characteristics,
phological insights into the sunfishes, the outstanding which we consider to be generally established at 6570
tasks of formally linking the genetic Mola sp. A clade cm TL, include clavus fin ray numbers, clavus ossicle
with nomenclature, and clarifying the status of the two numbers, state of the paraxial ossicle(s) and the pres-
Mola sp. B clades (Atlantic vs Pacific), currently limit ence of a smooth band back-fold and associated clavus
us to an interim update of the Fraser-Brunner (1951) edge indent in M. tecta. Also, the scale morphology
appears to be fully developed for M. tecta, while further Review of nominal Mola species
research is needed for the other Mola species. To confirm whether a scientific name consistent with
Interim key to the species of adult form Mola [modified the M.tecta characteristics was proposed in the past,
from Fraser-Brunner (1951)]: we reviewed the original descriptions of nominal Mola
species and other relevant names (Table2), as well
1.Smooth band (pre-clavus) with pronounced back- as associated historical literature. Due to the current
fold; clavus supported by 1517 rays [1315 prin- uncertainty in regards to linking the genetic clades of
cipal, 2 minor], 57 of the principal rays bearing Mola sp. Aand Mola sp. B with established nomencla-
ossicles; osscicles borne on paraxial rays sepa- ture, we did not attempt to assign species names to the
rate, of similar size to neighbouring; clavus edge nominal species descriptions, instead, we focused on
rounded, with an indent at the smooth band back- establishing whether any nominal species description
fold; without a protruding snout, head bump or matched M.tecta. In support of this evaluation, early
chin bump; short dorsolateral ridge to posterior Molidae taxonomic history was also reviewed.
end of pectoral fin; conical body scales of varying The original description of M. mola (T. mola, type
sizes, without branching tips, sparsely set locality in the Mediterranean) was brief, and included
..Mola tecta sp.nov. references to sunfish descriptions by Artedi (1738:
Smooth band without pronounced back-fold, or 61, 83), Bianchi (1746: 297) and Gronovius (1754: 55),
with a faint, superficial back-fold; other shape of who in turn referenced various early, often nebulous,
clavus; 8 or more ossicles.2 sunfish descriptions. Most of these were from the
2.Smooth band (pre-clavus) subtle, or moderately Mediterranean and the European seas, some describ-
pronounced; clavus supported by 1417 rays, 815 ing and illustrating what we now recognize as Mola
of which bear ossicles; clavus edge rounded; pro- forms (e.g. Rondelet, 1554: 424; Gessner, 1560: 158,
truding snout in some individuals; head bump, chin 159; Aldrovandi, 1613: 412), others Ranzania forms
bump, swollen dorso- and ventrolateral ridges in [e.g. Aldrovandi, 1613: 413; Jonstonus, 1657: Pl. 9 (No.
large individuals; body scales with slightly raised 1)]. As Linnaeus referenced both forms, it is not clear
rectangular midpoints in horizontal rows which sunfish he intended to describe. Although there
..Mola sp. A(presumably Mola ramsayi) is no known extant holotype for M. mola (sensu Parenti,
Smooth band pronounced; clavus supported by 2003), we know this species today from the historical
1013 rays, 89 of which bear ossicles; ossicles application of the name. Linnaeus Mola/Ranzania
invested with cuticle, which grows to form lobes ambiguity was noted by several authors (e.g. Retzius,
with fish growth, creating a wavy clavus edge; coni- 1785; Jacob, 1826; Steenstrup & Ltken, 1898) and
cal body scales with branching tips, densely set; pro- caused confusion in early sunfish systematics; how-
truding snout in some individuals; without or with ever, the name eventually came to mean M. mola as
moderate head bump; without or with small chin described in Fraser-Brunner (1951).
bump; swollen dorso- and ventrolateral ridges in Early sunfish systematics was also influenced by
large individuals differing opinions on how to resolve the taxonomy,
..Mola sp. B (presumably Mola mola) in particular for the comparatively minute pre-juve-
niles, the elongated Ranzania shapes, and the many
Source information for Mola sp. Aand Mola B mor- different sizes and shapes of Mola specimens. The
phology: head bump/chin bump: Yoshita etal. collective endeavours of many naturalists resulted
(2009), Matsuura (2015: fig.6), Sawai etal. (2015), in numerous proposed classifications, as well as an
Ahuir-Baraja etal. (2017), Sawai etal. (2017); pro- extensive web of references and synonymies, with sev-
truding snout: Fraser-Brunner (1951), Sawai E and eral nominal names based on re-grouping of existing
Nyegaard M (unpublished data); dorso- and ventro- species (Table2). The distinct pre-juveniles were ini-
lateral ridges: Thys etal. (2013: fig.1B), Matsuura tially viewed as a separate species, or genus, resulting
(2015: fig.6), Sawai E and Nyegaard M (unpublished in a number of nominal species names, all later syn-
data), smooth band: Yoshita etal. (2009), Sawai onymized with M.mola (Fraser-Brunner, 1951; Parenti,
etal. (2015), Ahuir-Baraja etal. (2017); smooth band 2003; Eschmeyer etal., 2017). In our view, the current
back-folds: M. Nyegaard and E. Sawai (unpublished lack of clarity on Mola spp. pre-juvenile morphology
data); clavus fin ray/ossicle counts: Yoshita etal. and identification precludes a conclusive assignment
(2009), Sawai (2016a), Sawai etal. (2017); clavus below genus level. However, these nominal species are
shape: Yoshita etal. (2009), Matsura (2015: fig.6), improbable candidates for M.tecta due to their type
Ahuir-Baraja etal. (2017: fig.1), Sawai etal. (2017); localities in the Mediterranean, USA and tropical
skin structure: Gaulide (1992), Sawai etal. (2015), seas (Table2). While a small number of pre-juvenile
Sawai (2016b). specimen localities are not known, they were unlikely
Figure 10. Locations of Mola tecta sp. nov.: holotype (black star); paratypes and other examined material (excluding
RMNH.PISC.D.2059 from the Dutch Coast) (grey circles); NNSW-1 (Yoshita et al., 2009), AY940816, AY940826 (Bass et al.,
2005) and one unnamed specimen from Gauldie (1992) (black squares).
to originate from the temperate Southern Hemisphere, O.mola [=M.mola], [and] belongs to the southern hem-
as this region was still relatively unexplored by sci- isphere but gave few taxonomic details. The specimen
ence at the time they were reported (Table2). became the name bearing type at the London Natural
Most of the nominal species names were based on History Museum (BMNH 1883.11.29.22). As the
larger typical sunfish specimens, with type localities Yoshita etal. (2009) Mola sp. C specimen was sampled
in the Mediterranean and European seas. Mola mola at the type locality of M.ramsayi in New South Wales,
was, until comparatively recently, considered the only and due to the unresolved discrepancy between the
Mola species present here; nevertheless, two records of morphology of Mola sp. Aand M.ramsayi, the question
Mola sp. Ain the Mediterranean (Ahuir-Baraja etal., arises whether Mola sp. C could be M.ramsayi, and
2017; fig.3) and one record of M.tecta from the Dutch Mola sp. Aanother species altogether? It is an unlikely
coast (see below under Broader Molidae literature) scenario however; photos of the newly restored M.ram-
preclude automatic assignment of all the European sayi holotype (Natural History Museum, London:
ocean sunfish names to M.mola. However, a careful www.instagram.com/p/BIzgLphgufI/) clearly show a
review of the nominal species descriptions, illustra- head bump and chin bump, a lack of, or inconspicuous,
tions and associated literature strongly suggests that smooth band, all inconsistent with the head profile
M.tecta was not among them (Table2). Of the nomi- and prominent smooth band of M.tecta.
nal species in Table2, three had type localities in the Orthagoriscus eurypterus Philippi, 1892 was
Southern Hemisphere, in areas we believe to be within described as a new species based on a large specimen
the core distribution of M.tecta, and warrant further (222cm TL, 247cm TBD), caught on 29 October 1889
discussion. off Chanaral in Chile. It was prepared and mounted
Aledon capensis Castelnau, 1861 was described prior to assessment, which likely rendered it some-
based on a 100cm TL specimen caught in Table what distorted (Schneider, 1930). The features, which
Bay near Cape Town, South Africa, where one of the made Philippi believe he had a new species, were su
M.tecta specimens in Bass etal. (2005) was also sam- forma mas alargada [the more elongated shape];
pled (Fig.10). However, A.capensis was trs-remar- la situacion de las aletas dorsal i anal, mucho mnos
quable par une avance en forme de nez, qui se prolonge atrasada; la dorsal principia casi en la mitad de la
au-dessus de la bouche [remarkable for a nose lonjitud del cuerpo [the position of the dorsal and
shape, extending beyond the mouth], and together anal fins with the dorsal fin sitting almost in the
with the description of a lobed clavus we conclude in middle of the body]; la forma mui ancha de ellas
agreement with Fraser-Brunner (1951), Parenti (2003) [the wideness of the fins], and the edge of the cla-
and Eschmeyer etal. (2017) that this specimen was vus, described as bien redondeado [well rounded]
M.mola (Table2). and no ofrece una sria de osificaciones cortantes i
Mola ramsayi (Giglioli, 1883) (originally separadas, parece mas bien todo osificado [not show-
Orthragoriscus ramsayi) was based on a large speci- ing a string of individual ossifications but seeming to
men [244cm TL according to Whitley (1931)] from be ossified in the entire length]. While the TBD:TL
New South Wales in Australia, exhibited at the 1883 ratio of 111% and the wide dorsal and anal fins are
International Fisheries Exhibition in the UK. Giglioli consistent with all three Mola species (Yoshita etal.,
(1883) examined it and concluded it differs from our 2009; this study), the combination of a rounded clavus
and an elongated body on such a large fish point to Eschmeyer etal. (2017), who list O.eurypterus as a
it possibly being M.tecta. The position of the dorsal junior synonym of M.ramsayi (Table2).
and anal fins almost in the middle of the body is
inconsistent with M.tecta, but could be an artefact
of preparation and mounting. The ossification of the Broader Molidae literature
entire length of the clavus edge, however, is incon- Within the broader Molidae literature, we did not
sistent with M.tecta, and we further note that while find any specimen descriptions matching M. tecta,
Philippi described the presence of a smooth band in with the exception of Lidth de Jeude (1890), who
detail, he did not make any mention of a back-fold. described and illustrated a smooth band parting
Phillipis specimen is extant at the National Museum off backwards from the prominent smooth band on
in Santiago de Chile (MNHNCH) as an unregis- a 223-cm-long female sunfish, stranded on the Dutch
tered specimen, which was not designated as holo- coast in December 1889. Lidth de Jeude could not find
type (Castro A, MNHNCH, personal communication, mention of a back-fold elsewhere in the literature,
November 2015). We examined this specimen via discussed the difficulty of species identification and
photographs provided to us by the museum, which tentatively settled on O. mola (=M. mola). His illus-
confirmed distortion in the specimen and loss of over tration later appeared in Martin & Drewry (1978:
half the clavus edge (Fig.11A). However, the skin fig. 171A) as the female form of M. mola, with a com-
structure is well preserved, is densely populated with ment that the back-fold was probably atypical. Lidth
scales (Fig.11B), resembles the skin of a mounted de Jeude's detailed and to-scale drawing (Fig. 12A)
Mola sp. B specimen described by Sawai etal. (2015: has a strong resemblance to M. tecta, and, while the
fig.3B) and is dissimilar to that of M.tecta with its TBD:TL ratio and count of eight ossicles are slightly
much wider set scales (Fig.9). Asmall stretch of the higher than our findings, his skin structure descrip-
upper clavus edge of the O.eurypterus specimen is tion is consistent with M. tecta. The mounted skin is
preserved (Fig.11A), and small ossicles are visible, extant at the Leiden Naturalis Biodiversity Center
interspersed with stretches of ossified clavus edge. in the Netherlands (RMNH.PISC.D.2059), and we
This corresponds to the description by Philippi examined this from photographs sent to us by the
(1892) and is consistent with the clavus morphology museum (Fig. 12BD); the body shape (Fig. 12C), skin
of M.mola. We conclude, in line with Anderssons structure (Fig. 12B) and smooth band with a back-
(1900) careful review of M.mola body dimensions, fold (Fig. 12D) suggest it is likely M. tecta. Juede did
that O.eurypterus is ett mycket stort exemplar af not describe his specimen as a new species, and with
den vanliga Orthagoriscus mola [a very large speci- no other indications of this species in the North East
men of the common O.mola (=M.mola)], and that Atlantic, where sunfish sightings and strandings are
the lack of clavus lobes was likely caused by damage not infrequent, it probably represents a rare occur-
during the preservation process. Our conclusion dif- rence of M. tecta outside the temperature waters of
fers from Fraser-Brunner (1951), Parenti (2003) and the Southern Hemisphere.
Figure11. Presumed Mola mola: (A) mounted skin of Orthagoriscus eurypterus (Philippi, 1892) extant at the National
History Museum in Santiago de Chile (MNHNCH unregistered, no type status). Black line indicates extent of clavus edge
loss; (B) macroscopic skin structure on the body (1cm scale bar). Photographs by Augusto Tomas Cornejo Castro, MNHNCH.
Figure 12. Presumed Mola tecta: (A) Reproduction from Lidth de Jeude (1890) of 223 cm total length female sunfish,
stranded on the Dutch coast in 1889, originally identified as Orthragoriscus mola (now M. mola), showing the smooth band
back-fold (black arrow) observed by Lidth de Jeude; (B) macroscopic skin structure on the body behind the pectoral fin; (C)
whole body view (white box indicates area of back-fold) and (D) smooth band (white arrow) and back-fold (black arrows) on
the mounted skin (RMNH.PISC.D.2059), extant at the Naturalis Biodiversity Center, The Netherlands. Photographs: Ronald
de Ruiter, RMNH.
CONCLUSION (Table 2) were based on specimens from the Southern

Hemisphere (i.e. Castalneu, 1861; Giglioli, 1883;
Our genetic and morphological analyses support the
Philippi, 1892). By the time sunfish were recorded by
preliminary findings of Yoshita etal. (2009), that Mola
the scientific community in Australia in the 1860s (e.g.
sp. C is a species distinct from Mola sp. Aand Mola
Royal Society of Tasmania, 1869) and 1870s in New
sp. B.We have reviewed the relevant taxonomic litera- Zealand (e.g. Hutton, 1872, 1873), the Molidae litera-
ture and conclude that this is a new sunfish species ture was already extensive and complicated, with a
not previously described. Our analysis and review also general consensus emerging that most of the proposed
show that the CO1 locus, used as the standard DNA Mola species were probably the same (i.e. Nardo, 1840;
barcoding for many taxon, can, like the D-loop, be used Steenstrup & Ltken, 1898). Later, even M. ramsayi
to identify individuals of the three Mola species. did not always gain support as a separate species,
It seems baffling that such a large fish has gone despite surviving the Fraser-Brunner (1951) review as
unnoticed by the scientific community despite the a Southern Hemisphere species. McCann (1961), for
extensive interest in sunfish taxonomy following, example, summarized and discussed all New Zealand
and even predating the time of Linnaeus. In addition sunfish stranding records as M. mola, without any
to the difficulty of studying these fish due to their mention of M. ramsayi. In recent years, it has become
large adult sizes, the explanation may partly lie in generally accepted that both M. mola and M. ramsayi
that early research efforts were concerned predomi- occur around New Zealand and southern Australia (e.g.
nantly with European and American sunfish, and that Gauldie, 1992; Hutchins, 2001; Bray, 2008; Swainston,
only three of the many nominal Mola species names 2011; Stewart & Struthers, 2015). Preliminary results
from our Molidae biopsy program in Australia and Historia Natural (Chile) and Ronald de Ruiter from
New Zealand suggests that Mola sp. B (=presumably Naturalis Biodiversity Center (The Netherlands)
M. mola) is a relatively rare visitor in those waters, as well as members of the public Dr Joana Browne,
whereas M. tecta is relatively common (M. Nyegaard, Ian Robertson, Roscoe Le Compte, Ken and Grace
unpublished data). With few local M. mola for com- Logan, Melanie Hiller and James McKibbin. ALS
parison, and in light of the long standing taxonomic acknowledges the great support from colleagues Carl
confusion, M. tecta in this region has presumably been Struthers (photography), Jeremy Barker (process-
mistaken for M. mola and/or M. ramsayi, allowing it to ing, preservation, X-rays), Salme Kortet (registration
hide in broad daylight. Its detection constitutes the and databasing) and Michelle Freeborn (illustrating
first proposed addition to the genus in 125 years since the holotype). The holotype was collected by New
Philippi (1892), and 130 years since the last valid Mola Zealand MPI Scientific Observer Scott Yeoman. The
species, M. ramsayi (Giglioli, 1883), was described. work was partly undertaken under animal ethics per-
The description herein of M.tecta sp. nov. clearly mits R2542/12 from Murdoch University. The project
shows it as a separate species of Mola, and also high- was supported by the Systematics Research Fund
lights areas of further research to establish the degree (Linnean Society of London and the Systematics
of natural variation across its growth spectrum, geo- Association), the Home for Innovative Researchers
graphical areas and between the sexes, to provide and Academic Knowledge Users HIRAKU (Hiroshima
sufficient information for a comprehensive Mola University) and the NZ National Institute of Water
identification key to be developed. The taxonomic and Atmospheric Research Ltd. Core Funded Coasts
review also highlights the need to formally link other & Oceans Programme 2: Biological Resources subcon-
genetic Mola spp. clades with nomenclature, and fur- tract for fundamental knowledge of marine fish bio-
ther describe their morphologies across the large size diversity with the Museum of New Zealand Te Papa
spectra. This would enhance our ability to differenti- Tongarewa.
ate the Mola species and thereby better understand
their zoogeographies, an important basis for gauging
vulnerability to by-catch in fisheries a potential key REFERENCES
threatening process to M.mola populations, at least
locally (Jing etal., 2015). Like the other two Mola spe- Ahuir-Baraja AE, Yamanoue Y, Kubicek L. 2017. First con-
cies, the eggs, pre-juveniles and juveniles of M.tecta firmed record of Mola sp. Ain the western Mediterranean
Sea: morphological, molecular and parasitological findings.
need to be located and verified through molecular
Journal of Fish Biology 90: 11331141.
methods to describe all life stages. Our results to date
Andersson LG. 1900. Ngra anteckningar om Orthagoriscus
are, however, sufficient to conclude that M.tecta should
mola (L). fversigt af Kongliga Vetenskaps-Akademiens
be considered a valid species in the genus Mola.
Frhandlingar 57: 603633.
Aldrovandi U. 1613. De piscibus libri V et De cetis lib. unus.
Bononiae: Dellagamba.
ACKNOWLEDGEMENTS Artedi P. 1738. Sueci Genera Piscium: in quibus systema totum
We are grateful for the generous help of Mike Yates, ichthyologiae proponitur cum classibus, ordinibus, generum
Steve Hall and fisheries observers from the Australian characteribus, specierum differentiis, observationibus pluri-
Fisheries Management Authority, James Andrew, mis. Redactis speciebus 242 ad genera 52. Ichthyologiae pars
Stephen Brouwer and fisheries observers from the III.; Sueci Descriptiones specierum piscium. quos vivos prae-
New Zealand Marine Ministry of Primary Industries, sertim disseguit et examinavit, inter quos primario pisces
regni sueciae facile omnes accuratissime describuntur cum
Cheryl Pullar from Department of Conservation
non paucis aliis exoticis. Ichthyologiae Pars V. In: Artedi
(New Zealand), Martin de Beer and Paul Davis from
P. ed. Ichthyologia sive opera omnia de piscibus scilicet:
Sanford Limited (New Zealand), Dr Jean McKinnon
Bibliotheca ichthyologica. Philosophia ichthyologica. Genera
and Tessa Mills (University of Otago, New Zealand),
piscium. Synonymia specierum. Descriptiones specierum.
Dean Stutter and Jeff Forman from National Omnia in hoc genere perfectiora quam antea ulla posthuma
Institute of Water and Atmospheric Research (New vindicavit, recognovit, coaptavit et editit Carolus Linnaeus.
Zealand), all natural history museums across New Lugduni Batavorum: Wishoff, Pars III: 1-85 + unpaginated
Zealand and Australia, in particular museum staff index; ParsV: 1-118 + unpaginated index.
Paul Scofield (Christchurch), Emma Burns (Otago), Ayres WO. 1859. On new fishes of the Californian coast.
Tom Trnski (Auckland), Belinda Bauer and Kathryn Proceedings of the California Academy of Natural Sciences
Medlock (Hobart), Mark McGrouther (Sydney), and 18581862 (Series 1) 2: 2532.
Dianne Bray and Martin Gomon from Museum Bailly N. 2015. WoRMS taxon details: Mola Koelreuter,
Victoria (Melbourne). We are also indebted to Augusto 1766. Available at: http://www.marinespecies.org/aphia.
Toms Cornejo Castro from Museo Nacional de php?p=taxdetails&id=126233 (accessed on 19 March 2017).
Barnard KH. 1935. Notes on South African marine fishes. Chang C-H, Shao K-T, Lin H-Y, Chiu Y-C, Lee M-Y, Liu SH,
Annals of the South African Museum 15: 645658. Lin PL. 2016. DNA barcodes of the native ray-finned fishes
Bass AL, Dewar H, Thys T, Streelman JT, Karl SA. 2005. in Taiwan. Molecular Ecology Resources 2016 17: 796805..
Evolutionary divergence among lineages of the ocean sunfish Charleton W. 1668. Onomasticon zoicon, plerorumque anima-
family, Molidae (Tetraodontiformes). Marine Biology 148: lium differentias & nomina propria pluribus linguis expon-
405414. ens. Cui accedunt mantissa anatomica; et quaedam de variis
Bianchi G. 1746. Jani Planci Ariminensis De Mola pisce fossilium generibus. Londini: Jacobum Allestry.
ad Josephum Montium bononiensem. De Bononiensi Cleland J. 1862. On the anatomy of the short sunfish
Scientiarum Et Artium Instituto Atque Academia (Orthragoriscus mola). Natural History Review (London) 2:
Commentarii 2: 297303. 170185.
Bianchi G. 1755. Jani Planci Ariminensis De Mola pisce. Cuvier G. 1798. Tableau lmentaire de lhistoire naturelle des
Epistola altera ad Josephum Montium bononiensem. De animaux. Paris: Baudouin.
Bononiensi Scientiarum Et Artium Instituto Atque Academia Cuvier G. 1817. Le rgne animal distribu daprs son organi-
Commentarii 3: 331334. sation, pour servir de base a lhistoire naturelle des animaux
Bloch ME. 1785. conomische Naturgeschichte der Fische et dintroduction a lanatomie compare, Tome II: les reptiles,
Deutschlands: mit sieben und dreyssig Kupfertafeln nach les poissons, les mollusques et les annlides. Paris: Deterville.
Originalen. Naturgeschichte der Auslndischen fische. Darriba D, Taboada GL, Doallo R, Posada D. 2012. jModel-
Mit sechs und dresissig ausgemalten kupfern nach origi- Test 2: more models, new heuristics and parallel computing.
nalen. Pt. 2, volume 2. Berlin: Auf Kosten des Verfassers, Nature Methods 9: 772.
und in Commission in der Buchhandlung der Realschule. Daubenton LJ-M. 1787. Histoire Naturelle. Tome Troisieme.
Bloch ME. 1786. Ichtyologie, ou Histoire naturelle, gnrale et Contenant les Poissons. In: Panckoucke [C-K] Encyclopdie
particulire des poissons: Avec des figures enlumines, dessi- mthodique. Paris, Panckoucke, 240241.
nes daprs nature. Part 3. Berlin: lAuteur, Franois de la Duhamel du Monceau HL. 1777. Trait gnral, des pesches,
Garde. et histoire des poissons quelles fournissent, tant pour la sub-
Bloch ME, Schneider JG. 1801. M. E.Blochii, Systema sistance des hommes, que pour plusieurs autres usages qui
Ichthyologiae iconibus cx illustratum. Post obitum auctoris ont rapport aux arts et au commerce, Vol. 3. Paris: Saillant &
opus inchoatum absolvit, correxit, interpolavit Jo. Gottlob Nyon; Veuve Desaint.
Schneider, Saxo. Berolini: Sanderiano. Eschmeyer WN, Fricke R, van der Laan R. 2017. Catalog
Boc A, Diallo AB, Makarenkov V. 2012. T-REX: a web server of fishes: genera, species, references. Available at: http://
for inferring, validating and visualizing phylogenetic trees researcharchive.calacademy.org/research/ichthyology/cata-
and networks. Nucleic Acids Research 40: 573579. log/fishcatmain.asp (accessed on 4 April 2016).
Bonaparte CL. 1832. Iconografia delle fauna italica per le Forsskl P. 1775. Descriptiones animalium avium, amphibio-
quattro classi degli animali vertebrati. Tomo III. Pesci. Rome: rum, piscium, insectorum, vermium; quae in itinere orientali
sulviucci. observavit. Hauniae: Mlleri.
Bonnaterre PJ. 1788. Tableau encyclopdique et mthodique Fraser-Brunner A. 1951. The ocean sunfishes (family
des trois rgnes de la nature: Ichthyologie. ddi et prsent a Molidae). Bulletin of the British Museum of (Natural History).
M.Necker, ministre dtat, & directeur gnral des Finances: Zoology 1: 87121.
vi. Paris: Panckoucke. Gauldie RW. 1992. Plywood structure and mineralization in
Borlase W. 1758. The natural history of Cornwall. The air, cli- the scales of the ocean sunfishes, Mola mola and M.ramsayi.
mate, waters, rivers, lakes, sea and tides; of the stones, semi- Tissue & Cell 24: 263266.
metals, metals, tin, and the manner of mining; the constitution Gemmell NJ, Akiyama S. 1996. An efficient method for
of the stannaries, iron, copper, silver, lead, and gold, found in the extraction of DNA from vertebrate tissues. Trends in
Cornwall; vegetables, rare birds, fishes, shells, reptiles, and Genetics: TIG 12: 338339.
quadrupeds of the inhabitants, their manners, customs, plays Gessner C. 1560. Nomenclator aquatilium animalium. Icones
or interludes, exercises, and festivals; the Cornish language, animalium aquatilium in mari et dulcibus aquis degentium,
trade, tenures, and arts. Oxford: W. Jackson. plusquam DCC.cum nomenclaturis singulorum Latinis,
Boussuet F. 1558. De natura aquatilium carmen. Lugduni: Grecis, Italicis, Hispanicis, Gallicis, Germanicis, Anglicis, ali-
Matthiam Bonhome. jsq; interdum, per certos ordines digest. Tiguri: Excudebat
Bray DJ. 2008. Family Molidae. In: Gomon MF, Bray DJ, Christoph. Froschoverus.
Kuiter RH, eds. Fishes of Australias southern coast. Sydney: Giglioli HH. 1883. Zoology at the fisheries exhibition. II.
Reed New Holland, 858861. Notes on the Vertebrata. Nature (London) 28: 313316.
Cadenat J. 1959. Notes dichtyologie ouest-africaine: 24. Gill T. 1884. Synopsis of the plectognath fishes. Proceedings of
Molidae ouest-africains avec description dune espce the United States National Museum 7: 411427.
nouvelle: Pseudomola lassarati de Cte dIvoire. Bulletin Glover CJM. 1994. Molidae. In: Gomon MF, Glover JCM,
de lInstitut Franais dAfrique Noire Srie A, Sciences Kuiter RH, eds. The fishes of Australias South Coast.
Naturelles 21: 11121122. Adelaide: State Print, 915920.
Castelnau F. 1861. Memoire sur les poissons de lAfrique aus- Gmelin JF. 1788. Caroli a Linn Systema naturae per regna
trale. Paris: J.-B. Baillire et Fils. tria naturae: secundum classes, ordines, genera, species, cum
characteribus, differentiis, synonymis, locis. Tom. 1 Pars 3. Katayama E, Matsuura K. 2016. Fine structures of scales
Lipsiae: Beer. of ocean sunfishes (Actinopterygii, Tetraodontiformes,
Gray JE. 1854. Catalogue of fish collected and described by Molidae): another morphological character supporting phy-
Laurence Theodore Gronow, now in the British Museum. logenetic relationships of the molid genera. Bulletin of the
London: Woodfall & Kinder. National Museum of Nature and Science, Series A 42: 9598.
Gronovius LT. 1754. Museum ichthyologicum, sistens pis- Klein JT. 1742. Historiae piscium naturalis promovendae mis-
cium indigenorum & quorumdam exoticorum, qui in Museo sus tertius, De piscibus per branchias occultas spirantibus ad
Laurentii Theodori Gronovii, J.U. D.adservantur, descrip- justum numerum et ordinem redigendis, cum observationi-
tiones ordine systematico, accedunt nonnullorum exoticorum bus circa partes genitales rajae maris, et ovarium galei. Part
piscium icones aeri incisae. Lugduni Batavorum: Haak. 3, 48pp. Gedani: Litteris Schreiberianis, 1742.
Gronovius LT. 1763. Zoophylacii Gronoviani Fasciculus pri- Koelreuter IT. 1766. Piscium rariorum e mus. Petrop.
mus exhibens Animalia Quadrupeda, Amphibia atque Pisces, Exceptorum descriptiones continuatae. Novi Commentarii
quae in Museo suo adservat, rite examinavit, systematice, Academiae Scientiarum Imperialis Petropolitanae 10 pro
disposuit, descripsit, atque iconibus illustravit. Lugduni Anno MDCCLXIV 329351.
Batavorum: Haak. Kumar S, Stecher G, Tamura K. 2016. MEGA7: Molecular
Guindon S, Gascuel O. 2003. A simple, fast, and accurate Evolutionary Genetics Analysis Version 7.0 for bigger data-
algorithm to estimate large phylogenies by maximum likeli- sets. Molecular Biology and Evolution 33: 18701874.
hood. Systematic Biology 52: 696704. Lacpde BGE. 1798. Histoire naturelle des poisons. Tom.
Gnther A. 1870. Catalogue of the fishes in the British Premie. Paris: Plassan.
Museum. Volume 8.Catalogue of the Physostomi, contain- Larochelle CR, Pickens FAT, Burns MD, Sidlauskas BL.
ing the families Gymnotidae, Symbranchidae, Muraenidae, 2016. Long-term isopropanol storage does not alter fish mor-
Pegasidae, and of the Lophopranchii, Plectognathi, Dipnoi, phometrics. Copeia 104: 411420.
Ganoidei, Chondropterygii, Cyclostomata, Leptocardii, in the Lidth de Jeude, Th. W. van L. 1890. On a large specimen of
British Museum. London: Taylor & Francis. Orthragoriscus on the Dutch coast. Notes from the Leyden
Harting P. 1865. Notices zoologiques, anatomiques et his- Museum 12: 189195.
tologiques, sur lOrthragoriscus ozodura; suivies de considra- Linard E. 1840. Description dune nouvelle espce du genre
tions sur lostogense des tlostieus en general. Amsterdam: mole (Orthagoriscus, Schn.) dcouverte lle Maurice. Revue
CG Vanderpost. Zoologique par la Socit Cuvierienne 3: 291292.
Hasegawa M, Kishino H, Yano T. 1985. Dating of the Linnaeus C. 1758. Systema naturae per regna tria naturae,
human-ape splitting by a molecular clock of mitochondrial secundum classes, ordines, genera, species, cum characteri-
DNA. Journal of Molecular Evolution 22: 160174. bus, differentiis, synonymis, locis. Tomus I.Editio decima,
Heemstra PC. 1986. Family Molidae. In: Smith MM, Heemstra reformata. Holmiae: Laurentii Salvius.
PC, eds. Smiths sea fishes, 6th edn. Berlin: Springer, 907908. Linnaeus C. 1766. Systema naturae per regna tria naturae,
Hutchins JB. 2001. Molidae. Molas (ocean sunfishes). In: secundum classes, ordines, genera, species, cum characteri-
Carpenter KE, Niehm VH, eds. FAO species identification bus, differentiis, synonymis, locis. Tomus I.Editio duodecima,
guide for fishery purposes. The living marine resources of reformata. Holmiae: Salvius.
the Western Central Pacific. Volume 6.Bony fishes part Martin FD, Drewry GE. 1978. Development of fishes of the
4 (Labridae to Latimeriidae), estuarine crocodiles, sea Mid-Atlantic Bight: an atlas of egg, larval and juvenile stages.
turtles, sea snakes and marine mammals. Rome: Food Volume 6: Stomateidae through Ogcocephalidae. Fort Collins:
and Agriculture Organization of the United Nations, US Fish and Wildlife Service, US Department of the Interior.
39663968. Matsuura K. 2015. Taxonomy and systematics of tetraodon-
Hutton FW. 1872. Catalogue, with diagnoses of the species. In: tiform fishes: a review focusing primarily on progress in
Hutton J, Hector FW, eds. Fishes of New Zealand. Wellington: the period from 1980 to 2014. Ichthyological Research 62:
James Hughes Printer, 193. 72113.
Hutton FW. 1873. Atr. XXVIII. Contributions to New Zealand McCann C. 1961. The sunfish, Mola mola (L.) in New Zealand
ichthyology. Transactions and Proceedings of the New waters. Records of the Dominion Museum 4: 720.
Zealand Institute 5: 259272. Mitchill SL. 1828. Description of an apparently new species
Jacob A. 1826. Observations respecting the Diodon or Tetrodon of Diodon. Annals of the Lyceum of Natural History of New
mola, or short Sun-Fish of Pennant. Dublin Philosophical York 2: 264265.
Journal and Scientific Review 6: 443448. Nakamura I, Goto Y, Sato K. 2015. Ocean sunfish rewarm at
Jing L, Zapfe G, Shao K-T, Leis JL, Matsuura K, Hardy the surface after deep excursions to forage for siphonophores.
G, Liu M, Robertson R, Tyler J. 2015. The IUCN Red The Journal of Animal Ecology 84: 590603.
List of Threatened Species 2015: Mola mola. Available Nardo GD. 1827a. Prodromus observationum et disquisitio-
at: http://dx.doi.org/10.2305/IUCN.UK.2015-4.RLTS. num Adriaticae ichthyologiae. Giornale di Fisica, Chimica,
T190422A1951231.en (accessed on 21 March 2017). Storia Naturale Medicina Ed Arti, Pavia Dec. II 10: 2240.
Jonstonus J. 1657. Historiae naturalis de quadrupetibus Nardo GD. 1827b. Estratto di una memoria ittiologica inedita.
libri: cum aeneis figuris; [Historiae naturalis de piscibus et Giornale di Fisica, Chimica, Storia Naturale Medicina Ed
cetis libri V]. Francofurti ad Moenum: Math. Meriani. Arti, Pavia Dec. II 10: 102105, 209213.
Nardo GD. 1840. Considerazioni sulla famiglia dei pesci Mola, Royal Society of Tasmania. 1869. List of objects presented
e sui caratteri che li distinguono. Annali delle Scienze del to the museum during 1869, with names of donors. Monthly
Regno Lombardo-Veneto 10: 105112. Notices of Papers and Proceedings of the Royal Society of
Pallas PS. 1770. Spicilegia zoologica quibus novae imprimis et Tasmania 8587.
obscurae animalium species iconibus, descriptionibus atque Ruysch F. 1710. Thesaurus animalium primus: cum figuris
commentariis illustrantur. Version 1.Fasciculus Octavus: aeneis. Het eerste cabinet der dieren. Amsteldami: Wolters.
154, 5 plt. Berolini: Gottlieb August Lange. Sabaj MH. 2016. Standard symbolic codes for institutional
Parenti P. 2003. Family Molidae Bonaparte 1832 molas or resource collections in herpetology and ichthyology: an Online
ocean sunfishes. California Academy of Sciences Annotated Reference. Version 6.5 [16 August 2016]. Available at: http://
Checklist of Fishes 18: 19. www.asih.org (accessed on 19 April 2017).
Pennant T. 1776. British zoology, Volume 3: ClassIII. Reptiles. Sagara K, Yoshita Y, Nishibori M, Kuniyoshi H, Umino T,
IV. Fish, 4th edn. London: Benj While. Sakai Y, Hashimoto H, Gushima K. 2005. Coexistence of
Philippi RA. 1892. Algunos peces de Chile. Las rayas, two clades of the ocean sunfish Mola mola (Molidae) around
Callorrhynchus i Orthagoriscus Chilenos. Anales del Museo the Japan cost. Japanese Journal of Ichthyology 52: 3539.
Nacional de Chile, Primera seccion, Zoolojia 3: 116. Salviani I. 1554. Aquatilium animalium historiae, liber pri-
Rafinesque CS. 1810a. Caratteri di alcuni nuovi generi e mus: cum eorumdem formis, aere excusis. Rom: Apud eun-
nuove specie di animali e piante della Sicilia, con varie dem Hippolytum Salvianum.
osservazioni sopra i medisimi. Palermo: Sanfilippo. Sawai E. 2016a. Morphological identifications of preserved
Rafinesque CS. 1810b. Indice dittiologia Siciliana, ossia, specimens of Mola in the Kagoshima University Museum.
Catalogo metodico dei nomi Latini, Italiani, e Siciliani Nature of Kagoshima 42: 343347.
dei pesci, che si rinvengono in Sicilia. Disposti secondo un Sawai E. 2016b. Validity and position of body scales as a taxo-
metodo naturale. eseguito da un appendice che contiene la nomic character in Mola sunfishes. Nature of Kagoshima 42:
descrizione di alcuni nuovi pesci siciliani. Illustrato da due 349352.
piance. Messina: Giovanni del Nobolo. Sawai E, Yamanoue Y, Mochizuki T, Sakai Y. 2015. Notes
Rambaut A, Suchard MA, Xie D, Drummond AJ. 2014. on the morphological characteristics of extremely large
Tracer v1.6. Available at: http://beast.bio.ed.ac.uk/Tracer. specimens of Mola sunfishes (Tetradontiformes, Molidae) in
Ranzani C. 1839. Dispositio familiae Molarum in genera et in Japanese museum collections. Bulletin of Ibaraki Natural
species. Novi Commentarii Academiae Scientiarum Instituti Museum 18: 6570.
Bononiensis 3: 6381. Sawai E, Yamanoue Y, Jawad L, Al-Mamry J, Sakai Y. 2017.
Redi F. 1684. Osservazioni intorno agli animali viventi che si Molecular and morphological identification of Mola sunfish
trovano negli animali viventi. Firenze: Piero Martini. specimens (Actinopterygii: Tetraodontiformes: Molidae) from
Retzius AJ. 1785. Tetrodon mola, beskrifven. Kongliga the Indian Ocean. Species Diversity 22: 99104.
Vetenskaps Academiens Nya Handlingar 6: 115121. Sawai E, Yamanoue Y, Sakai Y, Hashimoto H. 2009. On the
Retzius AJ. 1800. Faunae suecicae a Carolo Linn equ. abnormally morphological forms in Mola sunfish (Mola spp.
inchoatae pars prima: sistens Mammalia, Aves, Amphibia Aand B) taken from Japanese coastal waters. Journal of the
et Pisces Sueciae / quam recognovit, emendavit et auxit Graduate School of Biosphere Science, Hiroshima University
Andreas Joannes Retzius. Lipsiae: Lebrecht Crusium. 48: 917.
Risso A. 1810. Ichthyologie de Nice, ou, histoire naturelle Schneider CO. 1930. Algunas observaciones sobre el pez luna
des poissons du dpartement des Alpes maritimes. Paris: F. (*) (Mola Mola (Linn) Gilbert). Revista Chilena de Historia
Schoell. Natural 34: 200207.
Risso A. 1826. Histoire naturelle des principales productions Shaw G. 1804. General zoology or systematic natural history.
de lEurope mridionale, et particulirement de celles des Volume 5, Part 2: Pisces. With plates from the first authorities
environs de Nice et des Alpes maritimes. Tome Troisieme. and most select specimens engraved principally by Mr. Heath.
Paris: F. G. Levrault. London: G. Kearsley.
Rondelet G. 1554. Libri de Piscibus Marinis, in quibus verae Steenstrup JS, Ltken CF. 1898. Spolia Atlantica: bidrag
Piscium effigiens expressae sunt. Quae in tota Piscium his- til kundskab om klump- eller maanefiskene (Molidae). Det
toria contineantur, indicat Elenchus pagina nona et decima. Kongelige Danske videnskabernes selskabs skrifter, 6.Rkke,
Postremo accesserunt indices necessarii. Lugduni: Bonhomme. Naturvidenskabelig og mathematisk afdeling 9: 5102.
Rondelet G. 1558. La seconde partie de LHistoire entiere Stewart AL, Struthers CD. 2015. Family Molidae. In:
des poissons / compose premierement en latin par mais- Roberts CD, Stewart AL, Struthers CD, eds. The fishes of
tre Guilaume Rondelet ...; maintenant traduite en franois New Zealand. Wellington: Te Papa Press, 17451748.
[par Laurent Joubert] sans auoir rien omis estant necessaire Storer DH. 1839. A report on the fishes of Massachusetts.
lintelligence dicelle; avec leurs pourtraits au naif. Lyon: Boston Journal of Natural History 2: 289558.
Bonhome a la Masse DOr. Swainson W. 1838. On the natural history and classification
Ronquist F, Teslenko M, van der Mark P, Ayres DL, of fishes, amphibians, and reptiles, or monocardian animals,
Darling A, Hhna S, Larget B, Liu L, Suchard MA, Vol. 1. London: Longman, Orme, Brown, Green & Longmans.
Huelsenbeck JP. 2012. MrBayes 3.2: efficient Bayesian Swainson W. 1839. On the natural history and classification
phylogenetic inference and model choice across a large model of fishes, amphibians, & reptiles, or monocardian animals,
space. Systematic Biology 61: 539542. Vol. 2. London: Longman, Orme, Brown, Green & Longmans.
Swainston R. 2011. Swainstons fishes of Australia: The com- Transactions of the Royal Society of London, Series B:
plete illustrated guide. Hawthorn: Penguin Books Australia. Biological Sciences 360: 18471857.
Thys TM, Whitney J, Hearn A, Weng KC, Peaherrera C, Warnow T. 2012. Standard maximum likelihood analyses of
Jawad L, Alfaro-Shigueto J, Mangel JC, Karl SA. 2013. alignments with gaps can be statistically inconsistent. PLoS
First record of the southern ocean sunfish, Mola ramsayi, in Currents: Tree of Life 4: RRN1308.
the Galpagos Marine Reserve. Marine Biodiversity Records Whitley GP. 1931. Studies in ichthyology. No.4. Records of the
6: 14. Australian Museum 18: 96134, xixvi.
Thys T, Ryan JP, Weng KC, Erdmann M, Tresnati J. 2016. Willughby F. 1686. De historia piscium libri quatuor, jussu
Tracking a marine ecotourism star: movements of the short & sumptibus societatis regiae Londinensis editi. Totum opus
ocean sunfish Mola ramsayi in Nusa Penida, Bali, Indonesia. recognovit, coaptavit, supplevit, librum etiam primum &
Journal of Marine Biology 2016: 16. Article ID: 8750193. secundum integros adjecit J.Raius e societate regia. Oxonii:
Tyler JC. 1980. Osteology, phylogeny, and higher classification Sheldoniano.
of the fishes of the order Plectognathi (Tetraodontiformes). Yamanoue Y, Miya M, Matsuura K, Katoh M, Sakai H,
NOAA Technical Report NMFS Circular 434. Nishida M. 2004. Mitochondrial genomes and phylog-
Walbaum JJ. 1792. Petri Artedi sueci genera piscium in eny of the ocean sunfishes (Tetraodontiformes: Molidae).
quibus systema totum ichthyologiae proponitur cum clas- Ichthyological Research 51: 269273.
sibus, ordinibus, generum characteribus, specierum differ- Yoshita Y, Yamanoue Y, Sagara K, Nishibori M, Kuniyoshi
entiis, observationibus plurimis. Redactis Speciebus 242 H, Umino T, Sakai Y, Hashimoto H, Gushima K.
ad Genera 52. Ichthyologiae pars III. Grypeswaldiae: Ant. 2009. Phylogenetic relationship of two Mola sunfishes
Ferdin. Rose. (Tetraodontiformes: Molidae) occurring around the coast of
Ward RD, Zemlak TS, Innes BH, Last PR, Hebert PD. Japan, with notes on their geographical distribution and mor-
2005. DNA barcoding Australias fish species. Philosophical phological characteristics. Ichthyological Research 56: 232244.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the online version of this article at the publisher's web-site:
Figure S1. Coloration of live Mola tecta sp. nov. (A) Swimming near the surface in Reserva Marina Isla
Chaaral, Chile, 2015. Still from footage by Csar Villarroel (https://vimeo.com/129499857); (B) stranded and re-
floated in Port Phillip Bay, Victoria, 1979. Photograph by Martin Gomon, Museums Victoria, Australia; (C) and
(D) caught and released alive from tuna longlines off New Zealand, July 2013. Photographed by fisheries observer
Marli Dee, New Zealand Ministry of Primary Industries. (A) and (B) identified from morphology, including a
smooth band back-fold (white arrows); (C) and (D) identified genetically (D-loop).
Table S1. National Center for Biotechnology Information(NCBI) Accession numbers for mtDNA D-loop and
cytochrome c oxidase 1 (CO1) sequences analysed as part of this study. Institutional codes according to Sabaj
(2016), except ABTC (Australian Biological Tissue Collection) and OMNZ (Otago Museum).
Table S2. Whole or partial Molidae cytochrome c oxidase 1 (CO1) sequences lodged with Barcode of Life Data
Systems (BOLD) and/or the National Center for Biotechnology Information, as of 19 March2017.
Table S3. Morphometric measurements of the Mola tecta sp. nov. holotype (NMNZ P.057679) after initial freez-
ing and thawing (fresh state), and after preservation (1month in formalin, followed by 1month in an isopropyl
alcohol stepping process).

10 1093@Zoolinnean@Zlx040

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

10 1093@Zoolinnean@Zlx040

Uploaded by

Copyright:

Available Formats

Zoological Journal of the Linnean Society, 2017, XX, 128. With 12 figures.

Hiding in broad daylight: molecular and

INTRODUCTION ranging in life history stages from pre-juveniles (e.g.

topographies, with no differences in the relationships

Mola mola (Linnaeus,1758)

Diagnosis: Genus diagnosis, modified from Parenti Mola ramsayi (Giglioli,1883)

the English name, Hoodwinker, pertains to the

Habitat and distribution: Mola tecta has been

Biology: It attains at least 242cm TL, probably more.

Remarks: The clavus fin ray structure of NZ19

CONCLUSION (Table 2) were based on specimens from the Southern

You might also like